J.D. Lomas
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
40 records found
1
AI tools have entered design teams faster than the shared practices for using them. Inside design teams, every designer already uses AI in their own way, but what one designer learns never reaches the rest of the group. This thesis investigated how AI can be integrated into the early-stage design workflow of a professional team in a way that turns individual AI use into a shared, collaborative team capability.
The research followed a Research through Design approach, grounded in interviews, workflow analysis and low-fidelity AI probes with designers in the aviation industry. It reframed the original design brief twice: first, the workflow difficulty was not communication between disciplines but sensemaking under pressure; second, the real gap was not a lack of AI tools, but the absence of a shared structure for using them.
These findings led to MAIK (Modular AI Kit), a three-layer framework that treats AI use as a team practice rather than an individual skill. MAIK Cards give designers physical AI capabilities to browse, sort and debate as a team. The MAIK Canvas is a worksheet that a team fills in together to build reusable Claude Skills. Skills are then chained into longer workflows with a human review step between each. MAIK is framed as a learning infrastructure, not a production infrastructure: its purpose is to build a team's AI literacy, not to automate its output.
MAIK was evaluated with seven IDE Master's students and a designer working in the aviation industry. The thesis concludes that in organisations where designers already use AI, the decisive move is not to add one more tool. It is to build the shared structure that turns individual AI use into a collective team capability. ...
The research followed a Research through Design approach, grounded in interviews, workflow analysis and low-fidelity AI probes with designers in the aviation industry. It reframed the original design brief twice: first, the workflow difficulty was not communication between disciplines but sensemaking under pressure; second, the real gap was not a lack of AI tools, but the absence of a shared structure for using them.
These findings led to MAIK (Modular AI Kit), a three-layer framework that treats AI use as a team practice rather than an individual skill. MAIK Cards give designers physical AI capabilities to browse, sort and debate as a team. The MAIK Canvas is a worksheet that a team fills in together to build reusable Claude Skills. Skills are then chained into longer workflows with a human review step between each. MAIK is framed as a learning infrastructure, not a production infrastructure: its purpose is to build a team's AI literacy, not to automate its output.
MAIK was evaluated with seven IDE Master's students and a designer working in the aviation industry. The thesis concludes that in organisations where designers already use AI, the decisive move is not to add one more tool. It is to build the shared structure that turns individual AI use into a collective team capability. ...
AI tools have entered design teams faster than the shared practices for using them. Inside design teams, every designer already uses AI in their own way, but what one designer learns never reaches the rest of the group. This thesis investigated how AI can be integrated into the early-stage design workflow of a professional team in a way that turns individual AI use into a shared, collaborative team capability.
The research followed a Research through Design approach, grounded in interviews, workflow analysis and low-fidelity AI probes with designers in the aviation industry. It reframed the original design brief twice: first, the workflow difficulty was not communication between disciplines but sensemaking under pressure; second, the real gap was not a lack of AI tools, but the absence of a shared structure for using them.
These findings led to MAIK (Modular AI Kit), a three-layer framework that treats AI use as a team practice rather than an individual skill. MAIK Cards give designers physical AI capabilities to browse, sort and debate as a team. The MAIK Canvas is a worksheet that a team fills in together to build reusable Claude Skills. Skills are then chained into longer workflows with a human review step between each. MAIK is framed as a learning infrastructure, not a production infrastructure: its purpose is to build a team's AI literacy, not to automate its output.
MAIK was evaluated with seven IDE Master's students and a designer working in the aviation industry. The thesis concludes that in organisations where designers already use AI, the decisive move is not to add one more tool. It is to build the shared structure that turns individual AI use into a collective team capability.
The research followed a Research through Design approach, grounded in interviews, workflow analysis and low-fidelity AI probes with designers in the aviation industry. It reframed the original design brief twice: first, the workflow difficulty was not communication between disciplines but sensemaking under pressure; second, the real gap was not a lack of AI tools, but the absence of a shared structure for using them.
These findings led to MAIK (Modular AI Kit), a three-layer framework that treats AI use as a team practice rather than an individual skill. MAIK Cards give designers physical AI capabilities to browse, sort and debate as a team. The MAIK Canvas is a worksheet that a team fills in together to build reusable Claude Skills. Skills are then chained into longer workflows with a human review step between each. MAIK is framed as a learning infrastructure, not a production infrastructure: its purpose is to build a team's AI literacy, not to automate its output.
MAIK was evaluated with seven IDE Master's students and a designer working in the aviation industry. The thesis concludes that in organisations where designers already use AI, the decisive move is not to add one more tool. It is to build the shared structure that turns individual AI use into a collective team capability.
Futures of disruptive technologies
An AI-enhanced foresight tool for exploring the future of disruptive technologies
This graduation project investigates how policymakers and strategists can be supported in exploring long-term futures shaped by the convergence of disruptive technologies. The project was conducted in collaboration with Quantum Delta NL and its Centre for Quantum and Society (CQS), which develops tools to explore the societal implications of emerging technologies. In an environment where multiple technologies, such as artificial intelligence, biotechnology, robotics, climate technologies, neurotechnology, and quantum technologies, are converging and evolving simultaneously, traditional strategic foresight tools often struggle to remain adaptable, engaging, and contextually relevant.
Existing scenario-based approaches frequently become static, generalized, or quickly outdated, while the growing complexity of technological convergence makes it increasingly difficult for decision-makers to imagine plausible long-term futures. This research therefore addresses the question of how foresight tools can support structured exploration of complex futures without overwhelming participants, while still fostering meaningful strategic engagement.
To address these challenges, the project developed DeltAI, an AI-enhanced foresight tool designed for collaborative exploration. DeltAI combines established foresight methodologies with generative artificial intelligence, not to predict future outcomes but to structure conversations that help participants imagine possible futures and reflect on their implications for present-day decision-making.
The experience unfolds through a four-phase structure (Frame, Explore, Act, and Reflect) in which users collectively define a scenario direction, select disruptive technologies to examine, and adopt assigned perspectives that shape discussion. Through a backcasting process, teams move from a distant future scenario toward the present, analysing how developments could unfold over time before proposing strategic interventions. Generative AI is then used to simulate cascading consequences, including second- and third-order effects, revealing systemic implications and potential blind spots.
Methodologically, the research follows a design-driven approach centred on iterative prototyping. Across five prototype iterations, the tool was continuously refined through user testing sessions with stakeholders. Each iteration generated qualitative insights through observations, group reflections, and questionnaires, which informed subsequent design decisions. The development process also involved experimenting with the AI-supported prototyping techniques of ‘vibe coding, which allows rapid translation of design concepts into functional digital prototypes.
The primary contribution of this project lies not in developing a predictive model of the future, but in demonstrating how generative AI can be integrated into foresight practices in a human-centred way. DeltAI positions artificial intelligence as an imagination scaffold that expands the range of perspectives and consequences participants consider, while preserving human dialogue, interpretation, and judgement as the core drivers of sense-making. The research shows that carefully constrained AI systems can enhance engagement, adaptability, and imaginative capacity in foresight workshops, particularly when combined with physical interaction mechanisms, role-based perspectives, and structured facilitation.
Ultimately, the project contributes both a practical foresight tool and a set of design principles for AI-enhanced strategic foresight and vibe coding for design practice. These insights illustrate how AI can augment, rather than replace, collective reflection on uncertain futures, supporting policymakers and strategists in navigating increasingly complex technological landscapes. ...
Existing scenario-based approaches frequently become static, generalized, or quickly outdated, while the growing complexity of technological convergence makes it increasingly difficult for decision-makers to imagine plausible long-term futures. This research therefore addresses the question of how foresight tools can support structured exploration of complex futures without overwhelming participants, while still fostering meaningful strategic engagement.
To address these challenges, the project developed DeltAI, an AI-enhanced foresight tool designed for collaborative exploration. DeltAI combines established foresight methodologies with generative artificial intelligence, not to predict future outcomes but to structure conversations that help participants imagine possible futures and reflect on their implications for present-day decision-making.
The experience unfolds through a four-phase structure (Frame, Explore, Act, and Reflect) in which users collectively define a scenario direction, select disruptive technologies to examine, and adopt assigned perspectives that shape discussion. Through a backcasting process, teams move from a distant future scenario toward the present, analysing how developments could unfold over time before proposing strategic interventions. Generative AI is then used to simulate cascading consequences, including second- and third-order effects, revealing systemic implications and potential blind spots.
Methodologically, the research follows a design-driven approach centred on iterative prototyping. Across five prototype iterations, the tool was continuously refined through user testing sessions with stakeholders. Each iteration generated qualitative insights through observations, group reflections, and questionnaires, which informed subsequent design decisions. The development process also involved experimenting with the AI-supported prototyping techniques of ‘vibe coding, which allows rapid translation of design concepts into functional digital prototypes.
The primary contribution of this project lies not in developing a predictive model of the future, but in demonstrating how generative AI can be integrated into foresight practices in a human-centred way. DeltAI positions artificial intelligence as an imagination scaffold that expands the range of perspectives and consequences participants consider, while preserving human dialogue, interpretation, and judgement as the core drivers of sense-making. The research shows that carefully constrained AI systems can enhance engagement, adaptability, and imaginative capacity in foresight workshops, particularly when combined with physical interaction mechanisms, role-based perspectives, and structured facilitation.
Ultimately, the project contributes both a practical foresight tool and a set of design principles for AI-enhanced strategic foresight and vibe coding for design practice. These insights illustrate how AI can augment, rather than replace, collective reflection on uncertain futures, supporting policymakers and strategists in navigating increasingly complex technological landscapes. ...
This graduation project investigates how policymakers and strategists can be supported in exploring long-term futures shaped by the convergence of disruptive technologies. The project was conducted in collaboration with Quantum Delta NL and its Centre for Quantum and Society (CQS), which develops tools to explore the societal implications of emerging technologies. In an environment where multiple technologies, such as artificial intelligence, biotechnology, robotics, climate technologies, neurotechnology, and quantum technologies, are converging and evolving simultaneously, traditional strategic foresight tools often struggle to remain adaptable, engaging, and contextually relevant.
Existing scenario-based approaches frequently become static, generalized, or quickly outdated, while the growing complexity of technological convergence makes it increasingly difficult for decision-makers to imagine plausible long-term futures. This research therefore addresses the question of how foresight tools can support structured exploration of complex futures without overwhelming participants, while still fostering meaningful strategic engagement.
To address these challenges, the project developed DeltAI, an AI-enhanced foresight tool designed for collaborative exploration. DeltAI combines established foresight methodologies with generative artificial intelligence, not to predict future outcomes but to structure conversations that help participants imagine possible futures and reflect on their implications for present-day decision-making.
The experience unfolds through a four-phase structure (Frame, Explore, Act, and Reflect) in which users collectively define a scenario direction, select disruptive technologies to examine, and adopt assigned perspectives that shape discussion. Through a backcasting process, teams move from a distant future scenario toward the present, analysing how developments could unfold over time before proposing strategic interventions. Generative AI is then used to simulate cascading consequences, including second- and third-order effects, revealing systemic implications and potential blind spots.
Methodologically, the research follows a design-driven approach centred on iterative prototyping. Across five prototype iterations, the tool was continuously refined through user testing sessions with stakeholders. Each iteration generated qualitative insights through observations, group reflections, and questionnaires, which informed subsequent design decisions. The development process also involved experimenting with the AI-supported prototyping techniques of ‘vibe coding, which allows rapid translation of design concepts into functional digital prototypes.
The primary contribution of this project lies not in developing a predictive model of the future, but in demonstrating how generative AI can be integrated into foresight practices in a human-centred way. DeltAI positions artificial intelligence as an imagination scaffold that expands the range of perspectives and consequences participants consider, while preserving human dialogue, interpretation, and judgement as the core drivers of sense-making. The research shows that carefully constrained AI systems can enhance engagement, adaptability, and imaginative capacity in foresight workshops, particularly when combined with physical interaction mechanisms, role-based perspectives, and structured facilitation.
Ultimately, the project contributes both a practical foresight tool and a set of design principles for AI-enhanced strategic foresight and vibe coding for design practice. These insights illustrate how AI can augment, rather than replace, collective reflection on uncertain futures, supporting policymakers and strategists in navigating increasingly complex technological landscapes.
Existing scenario-based approaches frequently become static, generalized, or quickly outdated, while the growing complexity of technological convergence makes it increasingly difficult for decision-makers to imagine plausible long-term futures. This research therefore addresses the question of how foresight tools can support structured exploration of complex futures without overwhelming participants, while still fostering meaningful strategic engagement.
To address these challenges, the project developed DeltAI, an AI-enhanced foresight tool designed for collaborative exploration. DeltAI combines established foresight methodologies with generative artificial intelligence, not to predict future outcomes but to structure conversations that help participants imagine possible futures and reflect on their implications for present-day decision-making.
The experience unfolds through a four-phase structure (Frame, Explore, Act, and Reflect) in which users collectively define a scenario direction, select disruptive technologies to examine, and adopt assigned perspectives that shape discussion. Through a backcasting process, teams move from a distant future scenario toward the present, analysing how developments could unfold over time before proposing strategic interventions. Generative AI is then used to simulate cascading consequences, including second- and third-order effects, revealing systemic implications and potential blind spots.
Methodologically, the research follows a design-driven approach centred on iterative prototyping. Across five prototype iterations, the tool was continuously refined through user testing sessions with stakeholders. Each iteration generated qualitative insights through observations, group reflections, and questionnaires, which informed subsequent design decisions. The development process also involved experimenting with the AI-supported prototyping techniques of ‘vibe coding, which allows rapid translation of design concepts into functional digital prototypes.
The primary contribution of this project lies not in developing a predictive model of the future, but in demonstrating how generative AI can be integrated into foresight practices in a human-centred way. DeltAI positions artificial intelligence as an imagination scaffold that expands the range of perspectives and consequences participants consider, while preserving human dialogue, interpretation, and judgement as the core drivers of sense-making. The research shows that carefully constrained AI systems can enhance engagement, adaptability, and imaginative capacity in foresight workshops, particularly when combined with physical interaction mechanisms, role-based perspectives, and structured facilitation.
Ultimately, the project contributes both a practical foresight tool and a set of design principles for AI-enhanced strategic foresight and vibe coding for design practice. These insights illustrate how AI can augment, rather than replace, collective reflection on uncertain futures, supporting policymakers and strategists in navigating increasingly complex technological landscapes.
As Artificial Intelligence (AI) becomes a staple in modern classrooms, educators face a growing dilemma: how to effectively embrace personalization without compromising instructional control or inadvertently encouraging student academic dishonesty. This research addresses the current "black box" nature of AI in schools, where students often use tools like ChatGPT as a shortcut to bypass thinking, leaving teachers sidelined and unable to monitor genuine progress. To solve this, the thesis proposes a transition from a two-way (student-AI) interaction to a three-way collaboration involving the teacher, the student, and the AI.
Central to this research is the "Flight Simulator" metaphor, a conceptual framework designed to strike a balance between student autonomy and teacher oversight. In this model, the Student is the Pilot, operating in a safe "cockpit" where they can practice and make mistakes without real-world failure. The AI acts as the Co-Pilot, providing Socratic guidance and hints rather than direct answers. Crucially, the Teacher remains in the Control Tower, setting learning objectives and monitoring high-level data "radar" to intervene only when a student veers off course.
The development of the system followed a three-cycle research journey involving a "Think Tank" of educators and students. Cycle 1 (Exploration) revealed that teachers fear becoming "AI police" and that students often over-rely on AI, failing to spot misinformation. These insights led to a set of design requirements focusing on content control, process over answers, and differentiated instruction. Cycle 2 (Prototyping) tested the boundaries of surveillance, revealing that "Structured Autonomy"—where students have private chat spaces but teachers receive progress and mood analytics—was the most effective approach to maintaining trust while ensuring accountability.
The final output is Cubo, a web-based platform featuring two specialized interfaces. The Student Cockpit uses a game-like structure and interest-based metaphors (e.g., explaining physics through soccer) to build AI literacy. It forces students to engage in critical thinking, such as fact-checking AI claims, before completing a lesson. The Teacher Control Tower offers an automated dashboard that highlights struggling students and enables teachers to upload their own curriculum, ensuring the AI remains aligned with specific classroom objectives. To ease the digital transition, the system includes a physical "Starter Kit" to ground the technology in the real world.
The system was evaluated through a 10-day comparative study with 16 students. Results showed that while standard AI tools often decreased student confidence due to "hallucinations" and a lack of direction, the Cubo system significantly increased student confidence and improved their ability to spot misinformation (100% success rate). Teachers rated the system higher for ease of monitoring and felt empowered to return to their roles as mentors rather than administrators. Ultimately, this research demonstrates that AI, when d ...
Central to this research is the "Flight Simulator" metaphor, a conceptual framework designed to strike a balance between student autonomy and teacher oversight. In this model, the Student is the Pilot, operating in a safe "cockpit" where they can practice and make mistakes without real-world failure. The AI acts as the Co-Pilot, providing Socratic guidance and hints rather than direct answers. Crucially, the Teacher remains in the Control Tower, setting learning objectives and monitoring high-level data "radar" to intervene only when a student veers off course.
The development of the system followed a three-cycle research journey involving a "Think Tank" of educators and students. Cycle 1 (Exploration) revealed that teachers fear becoming "AI police" and that students often over-rely on AI, failing to spot misinformation. These insights led to a set of design requirements focusing on content control, process over answers, and differentiated instruction. Cycle 2 (Prototyping) tested the boundaries of surveillance, revealing that "Structured Autonomy"—where students have private chat spaces but teachers receive progress and mood analytics—was the most effective approach to maintaining trust while ensuring accountability.
The final output is Cubo, a web-based platform featuring two specialized interfaces. The Student Cockpit uses a game-like structure and interest-based metaphors (e.g., explaining physics through soccer) to build AI literacy. It forces students to engage in critical thinking, such as fact-checking AI claims, before completing a lesson. The Teacher Control Tower offers an automated dashboard that highlights struggling students and enables teachers to upload their own curriculum, ensuring the AI remains aligned with specific classroom objectives. To ease the digital transition, the system includes a physical "Starter Kit" to ground the technology in the real world.
The system was evaluated through a 10-day comparative study with 16 students. Results showed that while standard AI tools often decreased student confidence due to "hallucinations" and a lack of direction, the Cubo system significantly increased student confidence and improved their ability to spot misinformation (100% success rate). Teachers rated the system higher for ease of monitoring and felt empowered to return to their roles as mentors rather than administrators. Ultimately, this research demonstrates that AI, when d ...
As Artificial Intelligence (AI) becomes a staple in modern classrooms, educators face a growing dilemma: how to effectively embrace personalization without compromising instructional control or inadvertently encouraging student academic dishonesty. This research addresses the current "black box" nature of AI in schools, where students often use tools like ChatGPT as a shortcut to bypass thinking, leaving teachers sidelined and unable to monitor genuine progress. To solve this, the thesis proposes a transition from a two-way (student-AI) interaction to a three-way collaboration involving the teacher, the student, and the AI.
Central to this research is the "Flight Simulator" metaphor, a conceptual framework designed to strike a balance between student autonomy and teacher oversight. In this model, the Student is the Pilot, operating in a safe "cockpit" where they can practice and make mistakes without real-world failure. The AI acts as the Co-Pilot, providing Socratic guidance and hints rather than direct answers. Crucially, the Teacher remains in the Control Tower, setting learning objectives and monitoring high-level data "radar" to intervene only when a student veers off course.
The development of the system followed a three-cycle research journey involving a "Think Tank" of educators and students. Cycle 1 (Exploration) revealed that teachers fear becoming "AI police" and that students often over-rely on AI, failing to spot misinformation. These insights led to a set of design requirements focusing on content control, process over answers, and differentiated instruction. Cycle 2 (Prototyping) tested the boundaries of surveillance, revealing that "Structured Autonomy"—where students have private chat spaces but teachers receive progress and mood analytics—was the most effective approach to maintaining trust while ensuring accountability.
The final output is Cubo, a web-based platform featuring two specialized interfaces. The Student Cockpit uses a game-like structure and interest-based metaphors (e.g., explaining physics through soccer) to build AI literacy. It forces students to engage in critical thinking, such as fact-checking AI claims, before completing a lesson. The Teacher Control Tower offers an automated dashboard that highlights struggling students and enables teachers to upload their own curriculum, ensuring the AI remains aligned with specific classroom objectives. To ease the digital transition, the system includes a physical "Starter Kit" to ground the technology in the real world.
The system was evaluated through a 10-day comparative study with 16 students. Results showed that while standard AI tools often decreased student confidence due to "hallucinations" and a lack of direction, the Cubo system significantly increased student confidence and improved their ability to spot misinformation (100% success rate). Teachers rated the system higher for ease of monitoring and felt empowered to return to their roles as mentors rather than administrators. Ultimately, this research demonstrates that AI, when d
Central to this research is the "Flight Simulator" metaphor, a conceptual framework designed to strike a balance between student autonomy and teacher oversight. In this model, the Student is the Pilot, operating in a safe "cockpit" where they can practice and make mistakes without real-world failure. The AI acts as the Co-Pilot, providing Socratic guidance and hints rather than direct answers. Crucially, the Teacher remains in the Control Tower, setting learning objectives and monitoring high-level data "radar" to intervene only when a student veers off course.
The development of the system followed a three-cycle research journey involving a "Think Tank" of educators and students. Cycle 1 (Exploration) revealed that teachers fear becoming "AI police" and that students often over-rely on AI, failing to spot misinformation. These insights led to a set of design requirements focusing on content control, process over answers, and differentiated instruction. Cycle 2 (Prototyping) tested the boundaries of surveillance, revealing that "Structured Autonomy"—where students have private chat spaces but teachers receive progress and mood analytics—was the most effective approach to maintaining trust while ensuring accountability.
The final output is Cubo, a web-based platform featuring two specialized interfaces. The Student Cockpit uses a game-like structure and interest-based metaphors (e.g., explaining physics through soccer) to build AI literacy. It forces students to engage in critical thinking, such as fact-checking AI claims, before completing a lesson. The Teacher Control Tower offers an automated dashboard that highlights struggling students and enables teachers to upload their own curriculum, ensuring the AI remains aligned with specific classroom objectives. To ease the digital transition, the system includes a physical "Starter Kit" to ground the technology in the real world.
The system was evaluated through a 10-day comparative study with 16 students. Results showed that while standard AI tools often decreased student confidence due to "hallucinations" and a lack of direction, the Cubo system significantly increased student confidence and improved their ability to spot misinformation (100% success rate). Teachers rated the system higher for ease of monitoring and felt empowered to return to their roles as mentors rather than administrators. Ultimately, this research demonstrates that AI, when d
This thesis investigates the strategic integration of Artificial Intelligence (AI) to enhance user experience (UX) in the automotive sector, with a focus on Polestar, a premium electric vehicle manufacturer. Through an strategic and exploratory design-driven approach, the research identifies AI-powered UX opportunities that align with Polestar’s brand identity and user values. Methods included expert interviews, trend analysis, technology scouting, and co-design workshops. The study delivers four key outcomes: a value based future vision, an ideation tool in the form of a card deck, three prioritized opportunity tracks, and a strategic roadmap outlining design modules across implementation horizons. The identified opportunity tracks—Contextual Zone, Get to Know Each Other, and Confusion Remover are recommendations for prioritization in Polestar's research and development, and were mapped in the roadmap in form of design modules. Emphasizing context-awareness, trustworthiness, and joy of driving, the thesis provides actionable insights for Polestar’s UX team to guide AI-driven innovation and bring unique value to users and brand.
...
This thesis investigates the strategic integration of Artificial Intelligence (AI) to enhance user experience (UX) in the automotive sector, with a focus on Polestar, a premium electric vehicle manufacturer. Through an strategic and exploratory design-driven approach, the research identifies AI-powered UX opportunities that align with Polestar’s brand identity and user values. Methods included expert interviews, trend analysis, technology scouting, and co-design workshops. The study delivers four key outcomes: a value based future vision, an ideation tool in the form of a card deck, three prioritized opportunity tracks, and a strategic roadmap outlining design modules across implementation horizons. The identified opportunity tracks—Contextual Zone, Get to Know Each Other, and Confusion Remover are recommendations for prioritization in Polestar's research and development, and were mapped in the roadmap in form of design modules. Emphasizing context-awareness, trustworthiness, and joy of driving, the thesis provides actionable insights for Polestar’s UX team to guide AI-driven innovation and bring unique value to users and brand.
Adaptive Reuse of Water Heritage In The Dutch Polder Landscape
Bridging Theory and Practice in the Redevelopment of a Historic Pumping Station
This thesis focuses on the adaptive reuse of obsolete historic water infrastructure in the Dutch Polder landscape. It examines how these heritage structures can be reimagined to effectively serve contemporary social, cultural, and environmental needs. The goal of this research is to guide the development of an adaptive reuse project in this context. For this reason, the thesis consists of two parts: research on adaptive reuse of obsolete water infrastructure, and a practical design proposal that will apply the research findings on a real-world adaptive reuse project of a water heritage site: the pumping station Middelpolder in Amstelveen, the Netherlands.
First, the research uses a qualitative multi-method approach, by combining case studies, literature review, expert interviews, and public engagement. The framework by Vafaie, Remøy, and Gruis (2023), which defines ten success factors, is applied to perform an in-depth analysis of three existing reuse projects of water heritage in Dutch polders: DIJK, De Breek, and the Concertgemaal. Common elements across these cases include hybrid governance models, minimal and reversible architectural interventions, and a strong emphasis on public-facing (educational) programming. At the same time, each project dealt with challenges such as structural issues, financial feasibility, and regulatory complexity.
Secondly, this thesis further develops the concept for Atelier-Café Het Gemaal, a real-world reuse project in Amstelveen that integrates a café, creative workspace, and nature education hub in an obsolete water pumping station in the heart of the Middelpolder. This practical application is guided by the synthesis of patterns and insights from the data analysis. The project demonstrates a strategy of adaptive reuse that can reinforce cultural identity, local heritage, and promote community engagement while preserving the authenticity of a historic and monumental water pumping station.
To conclude, the research offers a series of reflections and recommendations for professionals in the fields of planning, heritage, and spatial development. The importance of phased implementation, platforming heritage as a design strategy, and building collaborations between public and private sectors are some examples of the recommendations. By connecting adaptive reuse to broader efforts in sustainability and circular economy, this research hopes to add a meaningful voice to ongoing academic conversations, as well as in the professional field of adaptive reuse of heritage buildings. ...
First, the research uses a qualitative multi-method approach, by combining case studies, literature review, expert interviews, and public engagement. The framework by Vafaie, Remøy, and Gruis (2023), which defines ten success factors, is applied to perform an in-depth analysis of three existing reuse projects of water heritage in Dutch polders: DIJK, De Breek, and the Concertgemaal. Common elements across these cases include hybrid governance models, minimal and reversible architectural interventions, and a strong emphasis on public-facing (educational) programming. At the same time, each project dealt with challenges such as structural issues, financial feasibility, and regulatory complexity.
Secondly, this thesis further develops the concept for Atelier-Café Het Gemaal, a real-world reuse project in Amstelveen that integrates a café, creative workspace, and nature education hub in an obsolete water pumping station in the heart of the Middelpolder. This practical application is guided by the synthesis of patterns and insights from the data analysis. The project demonstrates a strategy of adaptive reuse that can reinforce cultural identity, local heritage, and promote community engagement while preserving the authenticity of a historic and monumental water pumping station.
To conclude, the research offers a series of reflections and recommendations for professionals in the fields of planning, heritage, and spatial development. The importance of phased implementation, platforming heritage as a design strategy, and building collaborations between public and private sectors are some examples of the recommendations. By connecting adaptive reuse to broader efforts in sustainability and circular economy, this research hopes to add a meaningful voice to ongoing academic conversations, as well as in the professional field of adaptive reuse of heritage buildings. ...
This thesis focuses on the adaptive reuse of obsolete historic water infrastructure in the Dutch Polder landscape. It examines how these heritage structures can be reimagined to effectively serve contemporary social, cultural, and environmental needs. The goal of this research is to guide the development of an adaptive reuse project in this context. For this reason, the thesis consists of two parts: research on adaptive reuse of obsolete water infrastructure, and a practical design proposal that will apply the research findings on a real-world adaptive reuse project of a water heritage site: the pumping station Middelpolder in Amstelveen, the Netherlands.
First, the research uses a qualitative multi-method approach, by combining case studies, literature review, expert interviews, and public engagement. The framework by Vafaie, Remøy, and Gruis (2023), which defines ten success factors, is applied to perform an in-depth analysis of three existing reuse projects of water heritage in Dutch polders: DIJK, De Breek, and the Concertgemaal. Common elements across these cases include hybrid governance models, minimal and reversible architectural interventions, and a strong emphasis on public-facing (educational) programming. At the same time, each project dealt with challenges such as structural issues, financial feasibility, and regulatory complexity.
Secondly, this thesis further develops the concept for Atelier-Café Het Gemaal, a real-world reuse project in Amstelveen that integrates a café, creative workspace, and nature education hub in an obsolete water pumping station in the heart of the Middelpolder. This practical application is guided by the synthesis of patterns and insights from the data analysis. The project demonstrates a strategy of adaptive reuse that can reinforce cultural identity, local heritage, and promote community engagement while preserving the authenticity of a historic and monumental water pumping station.
To conclude, the research offers a series of reflections and recommendations for professionals in the fields of planning, heritage, and spatial development. The importance of phased implementation, platforming heritage as a design strategy, and building collaborations between public and private sectors are some examples of the recommendations. By connecting adaptive reuse to broader efforts in sustainability and circular economy, this research hopes to add a meaningful voice to ongoing academic conversations, as well as in the professional field of adaptive reuse of heritage buildings.
First, the research uses a qualitative multi-method approach, by combining case studies, literature review, expert interviews, and public engagement. The framework by Vafaie, Remøy, and Gruis (2023), which defines ten success factors, is applied to perform an in-depth analysis of three existing reuse projects of water heritage in Dutch polders: DIJK, De Breek, and the Concertgemaal. Common elements across these cases include hybrid governance models, minimal and reversible architectural interventions, and a strong emphasis on public-facing (educational) programming. At the same time, each project dealt with challenges such as structural issues, financial feasibility, and regulatory complexity.
Secondly, this thesis further develops the concept for Atelier-Café Het Gemaal, a real-world reuse project in Amstelveen that integrates a café, creative workspace, and nature education hub in an obsolete water pumping station in the heart of the Middelpolder. This practical application is guided by the synthesis of patterns and insights from the data analysis. The project demonstrates a strategy of adaptive reuse that can reinforce cultural identity, local heritage, and promote community engagement while preserving the authenticity of a historic and monumental water pumping station.
To conclude, the research offers a series of reflections and recommendations for professionals in the fields of planning, heritage, and spatial development. The importance of phased implementation, platforming heritage as a design strategy, and building collaborations between public and private sectors are some examples of the recommendations. By connecting adaptive reuse to broader efforts in sustainability and circular economy, this research hopes to add a meaningful voice to ongoing academic conversations, as well as in the professional field of adaptive reuse of heritage buildings.
Smart Kitchens and Smarter Conversations
Mapping Human-AI dynamics and expectations of applying GPT (Generative Pre-trained Transformer) models as an interface in smart kitchen ecosystems
We are on the verge of reversing a technological paradox. With the right steps, we can leverage the power of artificial intelligence to revolutionize the world of smart kitchen appliances. Imagine a kitchen where your fridge, oven, and cooking hob understand and anticipate your needs, making cooking a breeze and a joy. Generative AI, especially Large Language Models (LLMs), have given us a way to communicate naturally and humanly with machines. It’s almost like having a personal chef who knows exactly what you want, how you like it, and is ready to assist at any moment.
Now, think about your current smart kitchen appliances. They have amazing features, but how often do you find yourself frustrated by complicated settings or unintuitive controls? What if your kitchen as a whole, could not only understand your commands but also learn and adapt to your cooking habits? Let’s pose a question: “In the age of commercially available AI, what could we, as makers of kitchen appliances, do for you to make you choose our products and stay loyal to our brand?” Perhaps this next paraphrased question makes it more answerable – “For what job would you hire any of these appliances that we sell, to assist you in the kitchen?”
This project aims to kick off with the notion of getting GPTs to generate recipes and functions for the appliances. It studies consumers’ experiences modern kitchen appliances to understand what they truly need, want, and expect from a smart kitchen. It starts with a thorough exploration through literature reviews, online community trends, and direct user research to gather deep insights through experiments and interviews. These insights are then transformed into practical use cases for AI models. By crowdsourcing data, we create detailed user personas that showcase how personalized AI can elevate the cooking experience. We test the capabilities of OpenAI’s GPT-4 to design an actionable solution that ASKO can integrate into their future appliance lineup.
The result? A set of affordances of this proposed interface, a set of personas and a thorough understanding of market and customer value, complemented by a strategic story that outlines the logic necessary to deliver an extraordinary brand experience for ASKO consumers. Picture a kitchen where your oven knows the perfect temperature for your favourite dish, your fridge suggests recipes based on its contents, and your cooking hob adjusts its heat just the way you like it. This is not just a vision for the future—it’s a reality we are building today. By embracing this innovative approach, ASKO can lead the charge in transforming kitchens into intelligent, responsive, and delightful spaces, ensuring customer satisfaction and unwavering brand loyalty.
...
Now, think about your current smart kitchen appliances. They have amazing features, but how often do you find yourself frustrated by complicated settings or unintuitive controls? What if your kitchen as a whole, could not only understand your commands but also learn and adapt to your cooking habits? Let’s pose a question: “In the age of commercially available AI, what could we, as makers of kitchen appliances, do for you to make you choose our products and stay loyal to our brand?” Perhaps this next paraphrased question makes it more answerable – “For what job would you hire any of these appliances that we sell, to assist you in the kitchen?”
This project aims to kick off with the notion of getting GPTs to generate recipes and functions for the appliances. It studies consumers’ experiences modern kitchen appliances to understand what they truly need, want, and expect from a smart kitchen. It starts with a thorough exploration through literature reviews, online community trends, and direct user research to gather deep insights through experiments and interviews. These insights are then transformed into practical use cases for AI models. By crowdsourcing data, we create detailed user personas that showcase how personalized AI can elevate the cooking experience. We test the capabilities of OpenAI’s GPT-4 to design an actionable solution that ASKO can integrate into their future appliance lineup.
The result? A set of affordances of this proposed interface, a set of personas and a thorough understanding of market and customer value, complemented by a strategic story that outlines the logic necessary to deliver an extraordinary brand experience for ASKO consumers. Picture a kitchen where your oven knows the perfect temperature for your favourite dish, your fridge suggests recipes based on its contents, and your cooking hob adjusts its heat just the way you like it. This is not just a vision for the future—it’s a reality we are building today. By embracing this innovative approach, ASKO can lead the charge in transforming kitchens into intelligent, responsive, and delightful spaces, ensuring customer satisfaction and unwavering brand loyalty.
...
We are on the verge of reversing a technological paradox. With the right steps, we can leverage the power of artificial intelligence to revolutionize the world of smart kitchen appliances. Imagine a kitchen where your fridge, oven, and cooking hob understand and anticipate your needs, making cooking a breeze and a joy. Generative AI, especially Large Language Models (LLMs), have given us a way to communicate naturally and humanly with machines. It’s almost like having a personal chef who knows exactly what you want, how you like it, and is ready to assist at any moment.
Now, think about your current smart kitchen appliances. They have amazing features, but how often do you find yourself frustrated by complicated settings or unintuitive controls? What if your kitchen as a whole, could not only understand your commands but also learn and adapt to your cooking habits? Let’s pose a question: “In the age of commercially available AI, what could we, as makers of kitchen appliances, do for you to make you choose our products and stay loyal to our brand?” Perhaps this next paraphrased question makes it more answerable – “For what job would you hire any of these appliances that we sell, to assist you in the kitchen?”
This project aims to kick off with the notion of getting GPTs to generate recipes and functions for the appliances. It studies consumers’ experiences modern kitchen appliances to understand what they truly need, want, and expect from a smart kitchen. It starts with a thorough exploration through literature reviews, online community trends, and direct user research to gather deep insights through experiments and interviews. These insights are then transformed into practical use cases for AI models. By crowdsourcing data, we create detailed user personas that showcase how personalized AI can elevate the cooking experience. We test the capabilities of OpenAI’s GPT-4 to design an actionable solution that ASKO can integrate into their future appliance lineup.
The result? A set of affordances of this proposed interface, a set of personas and a thorough understanding of market and customer value, complemented by a strategic story that outlines the logic necessary to deliver an extraordinary brand experience for ASKO consumers. Picture a kitchen where your oven knows the perfect temperature for your favourite dish, your fridge suggests recipes based on its contents, and your cooking hob adjusts its heat just the way you like it. This is not just a vision for the future—it’s a reality we are building today. By embracing this innovative approach, ASKO can lead the charge in transforming kitchens into intelligent, responsive, and delightful spaces, ensuring customer satisfaction and unwavering brand loyalty.
Now, think about your current smart kitchen appliances. They have amazing features, but how often do you find yourself frustrated by complicated settings or unintuitive controls? What if your kitchen as a whole, could not only understand your commands but also learn and adapt to your cooking habits? Let’s pose a question: “In the age of commercially available AI, what could we, as makers of kitchen appliances, do for you to make you choose our products and stay loyal to our brand?” Perhaps this next paraphrased question makes it more answerable – “For what job would you hire any of these appliances that we sell, to assist you in the kitchen?”
This project aims to kick off with the notion of getting GPTs to generate recipes and functions for the appliances. It studies consumers’ experiences modern kitchen appliances to understand what they truly need, want, and expect from a smart kitchen. It starts with a thorough exploration through literature reviews, online community trends, and direct user research to gather deep insights through experiments and interviews. These insights are then transformed into practical use cases for AI models. By crowdsourcing data, we create detailed user personas that showcase how personalized AI can elevate the cooking experience. We test the capabilities of OpenAI’s GPT-4 to design an actionable solution that ASKO can integrate into their future appliance lineup.
The result? A set of affordances of this proposed interface, a set of personas and a thorough understanding of market and customer value, complemented by a strategic story that outlines the logic necessary to deliver an extraordinary brand experience for ASKO consumers. Picture a kitchen where your oven knows the perfect temperature for your favourite dish, your fridge suggests recipes based on its contents, and your cooking hob adjusts its heat just the way you like it. This is not just a vision for the future—it’s a reality we are building today. By embracing this innovative approach, ASKO can lead the charge in transforming kitchens into intelligent, responsive, and delightful spaces, ensuring customer satisfaction and unwavering brand loyalty.
Image-based AI for Industrial Design
How aligning semantics among desingers helps them use AI tools more effectively
This research explores the opportunities and limitations of image-based AI tools for industrial design. In collaboration with designers at Royal Gazelle, several tools were tested. AI tools can broaden and speed up the creative processes but also lack control for the user. AI models and designers use different terms to articulate perceptions and control output. This misalignment stems from misalignment among humans which translated to the training datasets of AI.
This report proposes co-creative image labelling sessions to align perception and articulation of perception in design teams. The aligned perception creates a strong vibe. Collectively labelling an image dataset increased alignment of the perception of images by 30.8% when comparing post- to pre-session questionnaire results. Additionally, participants used more of the same vocabulary after the session.
Training a low-rank adaptation model on these labelled datasets could externalise tacit design knowledge and can be used for further development on AI models that align with designers. ...
This report proposes co-creative image labelling sessions to align perception and articulation of perception in design teams. The aligned perception creates a strong vibe. Collectively labelling an image dataset increased alignment of the perception of images by 30.8% when comparing post- to pre-session questionnaire results. Additionally, participants used more of the same vocabulary after the session.
Training a low-rank adaptation model on these labelled datasets could externalise tacit design knowledge and can be used for further development on AI models that align with designers. ...
This research explores the opportunities and limitations of image-based AI tools for industrial design. In collaboration with designers at Royal Gazelle, several tools were tested. AI tools can broaden and speed up the creative processes but also lack control for the user. AI models and designers use different terms to articulate perceptions and control output. This misalignment stems from misalignment among humans which translated to the training datasets of AI.
This report proposes co-creative image labelling sessions to align perception and articulation of perception in design teams. The aligned perception creates a strong vibe. Collectively labelling an image dataset increased alignment of the perception of images by 30.8% when comparing post- to pre-session questionnaire results. Additionally, participants used more of the same vocabulary after the session.
Training a low-rank adaptation model on these labelled datasets could externalise tacit design knowledge and can be used for further development on AI models that align with designers.
This report proposes co-creative image labelling sessions to align perception and articulation of perception in design teams. The aligned perception creates a strong vibe. Collectively labelling an image dataset increased alignment of the perception of images by 30.8% when comparing post- to pre-session questionnaire results. Additionally, participants used more of the same vocabulary after the session.
Training a low-rank adaptation model on these labelled datasets could externalise tacit design knowledge and can be used for further development on AI models that align with designers.
How Might ChatGPT Improve the Accessibility of Quantum Computing?
A design research on improving Quantum Computing accessibility trough AI
The evolution of higher education and technological advancements has transformed the accessibility and understanding of complex subjects, such as quantum computing.
Traditionally, higher education has focused on specialized knowledge, making it challenging for individuals to explore diverse fields beyond their expertise. However, modern technologies have democratized access to information, enabling non-experts to engage with complex topics.
Quantum computing, grounded in the principles of quantum mechanics, is notoriously complex and has historically been accessible only to a select few. This exclusivity presents a challenge for those outside the field who wish to explore the possibilities of quantum technologies.
Recognizing this gap, the potential of ChatGPT to make quantum computing more approachable and understandable to non-experts can be explored. This exploration is not just about enhancing the functionality of ChatGPT but about envisioning a future where quantum computing is accessible to all, transforming the landscape of computation and innovation.
This research serves as an initial validation of the potential for large language models to act as an accessibility layer to quantum computing, having as a design tool the newly created Quantum Buddy 2.0, a custom GPT-4, fine-tuned through iterative feedback and testing.
A key part of the study is the development of a set of tasks designed to serve as benchmarks for evaluating the performance of large language models. Among these, one task was designed to enable those without a background in quantum mechanics to solve quantum encryption problems. Another task involved a test comparing Quantum Buddy in its previous versions with the standard model of ChatGPT. This comparison not only highlights the capabilities of these models but also provides insight for future evaluations. By utilizing these models, non-expert participants can now design and create with quantum computing, suggesting a future where quantum development moves forward rapidly.
The research thus demonstrates that ChatGPT significantly simplifies the process of programming quantum computers, facilitating learning and performance across a broad spectrum of abilities. It enables novices to engage in creative coding, allows coders to gain insights into quantum mechanics, and enhances the efficiency of quantum experts. This variety in application underscores ChatGPT’s role in democratizing access to quantum computing.
However, it is important to recognize that these findings represent the beginning of a much larger exploration. The evidence provided by the study points to a future where ChatGPT and similar large language models significantly enhance the non-expert's access to quantum computing. Despite the progress made, the journey is not without its challenges. The evolving capabilities of ChatGPT, along with the dynamic nature of programming languages and quantum computing itself, require ongoing adaptation and enhancement.
Overall, the research contributes to a foundational understanding of how large language models like ChatGPT can make quantum computing more accessible. This work lays the groundwork for further studies and developments, signaling a promising direction for the broader application and understanding of quantum computing technologies. ...
Traditionally, higher education has focused on specialized knowledge, making it challenging for individuals to explore diverse fields beyond their expertise. However, modern technologies have democratized access to information, enabling non-experts to engage with complex topics.
Quantum computing, grounded in the principles of quantum mechanics, is notoriously complex and has historically been accessible only to a select few. This exclusivity presents a challenge for those outside the field who wish to explore the possibilities of quantum technologies.
Recognizing this gap, the potential of ChatGPT to make quantum computing more approachable and understandable to non-experts can be explored. This exploration is not just about enhancing the functionality of ChatGPT but about envisioning a future where quantum computing is accessible to all, transforming the landscape of computation and innovation.
This research serves as an initial validation of the potential for large language models to act as an accessibility layer to quantum computing, having as a design tool the newly created Quantum Buddy 2.0, a custom GPT-4, fine-tuned through iterative feedback and testing.
A key part of the study is the development of a set of tasks designed to serve as benchmarks for evaluating the performance of large language models. Among these, one task was designed to enable those without a background in quantum mechanics to solve quantum encryption problems. Another task involved a test comparing Quantum Buddy in its previous versions with the standard model of ChatGPT. This comparison not only highlights the capabilities of these models but also provides insight for future evaluations. By utilizing these models, non-expert participants can now design and create with quantum computing, suggesting a future where quantum development moves forward rapidly.
The research thus demonstrates that ChatGPT significantly simplifies the process of programming quantum computers, facilitating learning and performance across a broad spectrum of abilities. It enables novices to engage in creative coding, allows coders to gain insights into quantum mechanics, and enhances the efficiency of quantum experts. This variety in application underscores ChatGPT’s role in democratizing access to quantum computing.
However, it is important to recognize that these findings represent the beginning of a much larger exploration. The evidence provided by the study points to a future where ChatGPT and similar large language models significantly enhance the non-expert's access to quantum computing. Despite the progress made, the journey is not without its challenges. The evolving capabilities of ChatGPT, along with the dynamic nature of programming languages and quantum computing itself, require ongoing adaptation and enhancement.
Overall, the research contributes to a foundational understanding of how large language models like ChatGPT can make quantum computing more accessible. This work lays the groundwork for further studies and developments, signaling a promising direction for the broader application and understanding of quantum computing technologies. ...
The evolution of higher education and technological advancements has transformed the accessibility and understanding of complex subjects, such as quantum computing.
Traditionally, higher education has focused on specialized knowledge, making it challenging for individuals to explore diverse fields beyond their expertise. However, modern technologies have democratized access to information, enabling non-experts to engage with complex topics.
Quantum computing, grounded in the principles of quantum mechanics, is notoriously complex and has historically been accessible only to a select few. This exclusivity presents a challenge for those outside the field who wish to explore the possibilities of quantum technologies.
Recognizing this gap, the potential of ChatGPT to make quantum computing more approachable and understandable to non-experts can be explored. This exploration is not just about enhancing the functionality of ChatGPT but about envisioning a future where quantum computing is accessible to all, transforming the landscape of computation and innovation.
This research serves as an initial validation of the potential for large language models to act as an accessibility layer to quantum computing, having as a design tool the newly created Quantum Buddy 2.0, a custom GPT-4, fine-tuned through iterative feedback and testing.
A key part of the study is the development of a set of tasks designed to serve as benchmarks for evaluating the performance of large language models. Among these, one task was designed to enable those without a background in quantum mechanics to solve quantum encryption problems. Another task involved a test comparing Quantum Buddy in its previous versions with the standard model of ChatGPT. This comparison not only highlights the capabilities of these models but also provides insight for future evaluations. By utilizing these models, non-expert participants can now design and create with quantum computing, suggesting a future where quantum development moves forward rapidly.
The research thus demonstrates that ChatGPT significantly simplifies the process of programming quantum computers, facilitating learning and performance across a broad spectrum of abilities. It enables novices to engage in creative coding, allows coders to gain insights into quantum mechanics, and enhances the efficiency of quantum experts. This variety in application underscores ChatGPT’s role in democratizing access to quantum computing.
However, it is important to recognize that these findings represent the beginning of a much larger exploration. The evidence provided by the study points to a future where ChatGPT and similar large language models significantly enhance the non-expert's access to quantum computing. Despite the progress made, the journey is not without its challenges. The evolving capabilities of ChatGPT, along with the dynamic nature of programming languages and quantum computing itself, require ongoing adaptation and enhancement.
Overall, the research contributes to a foundational understanding of how large language models like ChatGPT can make quantum computing more accessible. This work lays the groundwork for further studies and developments, signaling a promising direction for the broader application and understanding of quantum computing technologies.
Traditionally, higher education has focused on specialized knowledge, making it challenging for individuals to explore diverse fields beyond their expertise. However, modern technologies have democratized access to information, enabling non-experts to engage with complex topics.
Quantum computing, grounded in the principles of quantum mechanics, is notoriously complex and has historically been accessible only to a select few. This exclusivity presents a challenge for those outside the field who wish to explore the possibilities of quantum technologies.
Recognizing this gap, the potential of ChatGPT to make quantum computing more approachable and understandable to non-experts can be explored. This exploration is not just about enhancing the functionality of ChatGPT but about envisioning a future where quantum computing is accessible to all, transforming the landscape of computation and innovation.
This research serves as an initial validation of the potential for large language models to act as an accessibility layer to quantum computing, having as a design tool the newly created Quantum Buddy 2.0, a custom GPT-4, fine-tuned through iterative feedback and testing.
A key part of the study is the development of a set of tasks designed to serve as benchmarks for evaluating the performance of large language models. Among these, one task was designed to enable those without a background in quantum mechanics to solve quantum encryption problems. Another task involved a test comparing Quantum Buddy in its previous versions with the standard model of ChatGPT. This comparison not only highlights the capabilities of these models but also provides insight for future evaluations. By utilizing these models, non-expert participants can now design and create with quantum computing, suggesting a future where quantum development moves forward rapidly.
The research thus demonstrates that ChatGPT significantly simplifies the process of programming quantum computers, facilitating learning and performance across a broad spectrum of abilities. It enables novices to engage in creative coding, allows coders to gain insights into quantum mechanics, and enhances the efficiency of quantum experts. This variety in application underscores ChatGPT’s role in democratizing access to quantum computing.
However, it is important to recognize that these findings represent the beginning of a much larger exploration. The evidence provided by the study points to a future where ChatGPT and similar large language models significantly enhance the non-expert's access to quantum computing. Despite the progress made, the journey is not without its challenges. The evolving capabilities of ChatGPT, along with the dynamic nature of programming languages and quantum computing itself, require ongoing adaptation and enhancement.
Overall, the research contributes to a foundational understanding of how large language models like ChatGPT can make quantum computing more accessible. This work lays the groundwork for further studies and developments, signaling a promising direction for the broader application and understanding of quantum computing technologies.
Society currently faces many complex and significant challenges, dilemmas that lack straightforward solutions and are deeply interconnected. These problems demand a transformative approach that involves reevaluating people’s values, beliefs, attitudes, and behavior. Amidst these ongoing challenges, there is a collective hint of pessimism regarding society’s perception to effectively overcome them.
To navigate this era of transformative change, it becomes increasingly critical to engage in dialogue and define future goals. However, the process of generating specific visions and objectives for the future currently lacks guidance. This project seeks to bridge this gap by integrating futuring methods, creativity, co-creation, and Artificial Intelligence (AI). AI has the ability to enhance the abilities of people in this process. Through this innovative approach, the aim is to empower collective envisioning and goal-setting, enabling individuals and organizations to actively explore and shape desirable futures. More positive images of the future can motivate the broader collective as well.
The initial phase of the project involved an extensive exploration of existing literature on co-creation, future visions, the impact of positive future visions, creativity, and AI. In-depth interviews provided valuable insights from each distinct aspect: creating art with AI, facilitating co-creation, and developing future visions. Furthermore, a cultural probe shed light on the aspects individuals require more guidance in when contemplating the future. The findings underscored the relative ease with which people envisioned negative futures, and potential future inventions, while demonstrating a lesser emphasis on the underlying rationale and societal value of these inventions. Additionally, the perceived difficulty level varied considerably, and the outcomes were diverse. In this regard, the utilization of tools can prove beneficial, facilitating smoother and more efficient conversations while helping individuals initiate and navigate discussions with greater ease.
The accumulated basis of knowledge formed a buildable concept, applying the structure of the creative facilitation phases: problem, idea, and solution finding. This buildable concept was tested through four short sessions. Adding and altering its basis to establish the Minimal Viable Method (MVM) and expand to longer sessions. All sessions were conducted in a real setting with a diverse range of topics relevant to the problem owner(s), allowing for continuous improvement and addressing specific research questions during each session. Participants actively engaged in providing feedback and evaluating their experience based on predefined design criteria.
The end result of this project was a Version 1 (V1) design of a toolkit to guide people through a co-creation session using AI to create future visions. The toolkit allows participants to focus on their specific goals and type of futures, it is designed with a facilitator in mind to guide the process. Although, the toolkit includes explanations and guiding signs usable by people without knowledge of these elements. The toolkit includes underlayments, a card deck; a professional and newbie version, different inspiration museum topic sets, and a more elaborate optional manual to guide new users using AI, creating future visions, and vision statements. The toolkit is available through the website: https://future-framing.nl.
This project adds to the current literature by providing a new method to envision the future in co-creation. Evaluation results from these sessions validate the positive impact of the method on participants through specific design criteria. The majority of participants expressed interest in joining another workshop. Additionally, it finds a general interest of people to acquiring skills to use AI. The design goal, to spark empowerment and ownership in participants while creating their own desired future vision in a co-creation session, is successfully implemented with this project. The findings suggest future development of the V1 toolkit and method can further improve experience and outcome. This correlates with future developments with AI, that shape both too.
...
To navigate this era of transformative change, it becomes increasingly critical to engage in dialogue and define future goals. However, the process of generating specific visions and objectives for the future currently lacks guidance. This project seeks to bridge this gap by integrating futuring methods, creativity, co-creation, and Artificial Intelligence (AI). AI has the ability to enhance the abilities of people in this process. Through this innovative approach, the aim is to empower collective envisioning and goal-setting, enabling individuals and organizations to actively explore and shape desirable futures. More positive images of the future can motivate the broader collective as well.
The initial phase of the project involved an extensive exploration of existing literature on co-creation, future visions, the impact of positive future visions, creativity, and AI. In-depth interviews provided valuable insights from each distinct aspect: creating art with AI, facilitating co-creation, and developing future visions. Furthermore, a cultural probe shed light on the aspects individuals require more guidance in when contemplating the future. The findings underscored the relative ease with which people envisioned negative futures, and potential future inventions, while demonstrating a lesser emphasis on the underlying rationale and societal value of these inventions. Additionally, the perceived difficulty level varied considerably, and the outcomes were diverse. In this regard, the utilization of tools can prove beneficial, facilitating smoother and more efficient conversations while helping individuals initiate and navigate discussions with greater ease.
The accumulated basis of knowledge formed a buildable concept, applying the structure of the creative facilitation phases: problem, idea, and solution finding. This buildable concept was tested through four short sessions. Adding and altering its basis to establish the Minimal Viable Method (MVM) and expand to longer sessions. All sessions were conducted in a real setting with a diverse range of topics relevant to the problem owner(s), allowing for continuous improvement and addressing specific research questions during each session. Participants actively engaged in providing feedback and evaluating their experience based on predefined design criteria.
The end result of this project was a Version 1 (V1) design of a toolkit to guide people through a co-creation session using AI to create future visions. The toolkit allows participants to focus on their specific goals and type of futures, it is designed with a facilitator in mind to guide the process. Although, the toolkit includes explanations and guiding signs usable by people without knowledge of these elements. The toolkit includes underlayments, a card deck; a professional and newbie version, different inspiration museum topic sets, and a more elaborate optional manual to guide new users using AI, creating future visions, and vision statements. The toolkit is available through the website: https://future-framing.nl.
This project adds to the current literature by providing a new method to envision the future in co-creation. Evaluation results from these sessions validate the positive impact of the method on participants through specific design criteria. The majority of participants expressed interest in joining another workshop. Additionally, it finds a general interest of people to acquiring skills to use AI. The design goal, to spark empowerment and ownership in participants while creating their own desired future vision in a co-creation session, is successfully implemented with this project. The findings suggest future development of the V1 toolkit and method can further improve experience and outcome. This correlates with future developments with AI, that shape both too.
...
Society currently faces many complex and significant challenges, dilemmas that lack straightforward solutions and are deeply interconnected. These problems demand a transformative approach that involves reevaluating people’s values, beliefs, attitudes, and behavior. Amidst these ongoing challenges, there is a collective hint of pessimism regarding society’s perception to effectively overcome them.
To navigate this era of transformative change, it becomes increasingly critical to engage in dialogue and define future goals. However, the process of generating specific visions and objectives for the future currently lacks guidance. This project seeks to bridge this gap by integrating futuring methods, creativity, co-creation, and Artificial Intelligence (AI). AI has the ability to enhance the abilities of people in this process. Through this innovative approach, the aim is to empower collective envisioning and goal-setting, enabling individuals and organizations to actively explore and shape desirable futures. More positive images of the future can motivate the broader collective as well.
The initial phase of the project involved an extensive exploration of existing literature on co-creation, future visions, the impact of positive future visions, creativity, and AI. In-depth interviews provided valuable insights from each distinct aspect: creating art with AI, facilitating co-creation, and developing future visions. Furthermore, a cultural probe shed light on the aspects individuals require more guidance in when contemplating the future. The findings underscored the relative ease with which people envisioned negative futures, and potential future inventions, while demonstrating a lesser emphasis on the underlying rationale and societal value of these inventions. Additionally, the perceived difficulty level varied considerably, and the outcomes were diverse. In this regard, the utilization of tools can prove beneficial, facilitating smoother and more efficient conversations while helping individuals initiate and navigate discussions with greater ease.
The accumulated basis of knowledge formed a buildable concept, applying the structure of the creative facilitation phases: problem, idea, and solution finding. This buildable concept was tested through four short sessions. Adding and altering its basis to establish the Minimal Viable Method (MVM) and expand to longer sessions. All sessions were conducted in a real setting with a diverse range of topics relevant to the problem owner(s), allowing for continuous improvement and addressing specific research questions during each session. Participants actively engaged in providing feedback and evaluating their experience based on predefined design criteria.
The end result of this project was a Version 1 (V1) design of a toolkit to guide people through a co-creation session using AI to create future visions. The toolkit allows participants to focus on their specific goals and type of futures, it is designed with a facilitator in mind to guide the process. Although, the toolkit includes explanations and guiding signs usable by people without knowledge of these elements. The toolkit includes underlayments, a card deck; a professional and newbie version, different inspiration museum topic sets, and a more elaborate optional manual to guide new users using AI, creating future visions, and vision statements. The toolkit is available through the website: https://future-framing.nl.
This project adds to the current literature by providing a new method to envision the future in co-creation. Evaluation results from these sessions validate the positive impact of the method on participants through specific design criteria. The majority of participants expressed interest in joining another workshop. Additionally, it finds a general interest of people to acquiring skills to use AI. The design goal, to spark empowerment and ownership in participants while creating their own desired future vision in a co-creation session, is successfully implemented with this project. The findings suggest future development of the V1 toolkit and method can further improve experience and outcome. This correlates with future developments with AI, that shape both too.
To navigate this era of transformative change, it becomes increasingly critical to engage in dialogue and define future goals. However, the process of generating specific visions and objectives for the future currently lacks guidance. This project seeks to bridge this gap by integrating futuring methods, creativity, co-creation, and Artificial Intelligence (AI). AI has the ability to enhance the abilities of people in this process. Through this innovative approach, the aim is to empower collective envisioning and goal-setting, enabling individuals and organizations to actively explore and shape desirable futures. More positive images of the future can motivate the broader collective as well.
The initial phase of the project involved an extensive exploration of existing literature on co-creation, future visions, the impact of positive future visions, creativity, and AI. In-depth interviews provided valuable insights from each distinct aspect: creating art with AI, facilitating co-creation, and developing future visions. Furthermore, a cultural probe shed light on the aspects individuals require more guidance in when contemplating the future. The findings underscored the relative ease with which people envisioned negative futures, and potential future inventions, while demonstrating a lesser emphasis on the underlying rationale and societal value of these inventions. Additionally, the perceived difficulty level varied considerably, and the outcomes were diverse. In this regard, the utilization of tools can prove beneficial, facilitating smoother and more efficient conversations while helping individuals initiate and navigate discussions with greater ease.
The accumulated basis of knowledge formed a buildable concept, applying the structure of the creative facilitation phases: problem, idea, and solution finding. This buildable concept was tested through four short sessions. Adding and altering its basis to establish the Minimal Viable Method (MVM) and expand to longer sessions. All sessions were conducted in a real setting with a diverse range of topics relevant to the problem owner(s), allowing for continuous improvement and addressing specific research questions during each session. Participants actively engaged in providing feedback and evaluating their experience based on predefined design criteria.
The end result of this project was a Version 1 (V1) design of a toolkit to guide people through a co-creation session using AI to create future visions. The toolkit allows participants to focus on their specific goals and type of futures, it is designed with a facilitator in mind to guide the process. Although, the toolkit includes explanations and guiding signs usable by people without knowledge of these elements. The toolkit includes underlayments, a card deck; a professional and newbie version, different inspiration museum topic sets, and a more elaborate optional manual to guide new users using AI, creating future visions, and vision statements. The toolkit is available through the website: https://future-framing.nl.
This project adds to the current literature by providing a new method to envision the future in co-creation. Evaluation results from these sessions validate the positive impact of the method on participants through specific design criteria. The majority of participants expressed interest in joining another workshop. Additionally, it finds a general interest of people to acquiring skills to use AI. The design goal, to spark empowerment and ownership in participants while creating their own desired future vision in a co-creation session, is successfully implemented with this project. The findings suggest future development of the V1 toolkit and method can further improve experience and outcome. This correlates with future developments with AI, that shape both too.
Generative AI has had a profound impact on various sectors since mid-2022. The rapid adoption by students poses challenges for student assessment and raises concerns about student development. This thesis delves into the subject and proposes a positive implementation in education through play-based learning. The project uses a human-centred iterative design approach, consisting of co-design workshops with children, teachers and expert interviews. The research shows that teaching children about AI and generative AI can be done from a young age using play-based learning.
The study captures the diverse perceptions of elementary school children and teachers regarding generative AI. Key findings highlight children’s concerns and needs with regards to generative AI. Cooperatively formulated design criteria point towards children’s main interests for products or services with generative AI or AI to have a play-based nature, foster creativity and inspiration, safeguard their privacy and security and help them achieve their goals.
The research in this thesis also explores educators’ perceived challenges and interests in classroom integration of generative AI. Key findings include the perceived necessity for both teachers and students to learn about generative AI, the interest across different levels of education and that the age deemed appropriate for students to be introduced to generative AI is often considered to be at the end of primary school or at the start of secondary school.
From the research several design principles are proposed to create an engaging, ethical and human-centred AI learning experience. To facilitate this experience and to spearhead the integration of generative AI into education, a company named “PLAI” is conceptualised. The company aims to implement generative AI in education in a positive way by providing learning material which stimulates engagement, creativity and social learning while safeguarding privacy and security by offering play-based and scaffolded learning in workshops. Future recommendations and plans include the development of a multi-modal generative AI model which can run locally on school servers and is alignable with their curriculum.
Limitations of this study include a lack of direct interaction and research with high school students, a need to explore text based generative AI interactions with students, and a need to assess generative AI’s long-term effects on student well-being. ...
The study captures the diverse perceptions of elementary school children and teachers regarding generative AI. Key findings highlight children’s concerns and needs with regards to generative AI. Cooperatively formulated design criteria point towards children’s main interests for products or services with generative AI or AI to have a play-based nature, foster creativity and inspiration, safeguard their privacy and security and help them achieve their goals.
The research in this thesis also explores educators’ perceived challenges and interests in classroom integration of generative AI. Key findings include the perceived necessity for both teachers and students to learn about generative AI, the interest across different levels of education and that the age deemed appropriate for students to be introduced to generative AI is often considered to be at the end of primary school or at the start of secondary school.
From the research several design principles are proposed to create an engaging, ethical and human-centred AI learning experience. To facilitate this experience and to spearhead the integration of generative AI into education, a company named “PLAI” is conceptualised. The company aims to implement generative AI in education in a positive way by providing learning material which stimulates engagement, creativity and social learning while safeguarding privacy and security by offering play-based and scaffolded learning in workshops. Future recommendations and plans include the development of a multi-modal generative AI model which can run locally on school servers and is alignable with their curriculum.
Limitations of this study include a lack of direct interaction and research with high school students, a need to explore text based generative AI interactions with students, and a need to assess generative AI’s long-term effects on student well-being. ...
Generative AI has had a profound impact on various sectors since mid-2022. The rapid adoption by students poses challenges for student assessment and raises concerns about student development. This thesis delves into the subject and proposes a positive implementation in education through play-based learning. The project uses a human-centred iterative design approach, consisting of co-design workshops with children, teachers and expert interviews. The research shows that teaching children about AI and generative AI can be done from a young age using play-based learning.
The study captures the diverse perceptions of elementary school children and teachers regarding generative AI. Key findings highlight children’s concerns and needs with regards to generative AI. Cooperatively formulated design criteria point towards children’s main interests for products or services with generative AI or AI to have a play-based nature, foster creativity and inspiration, safeguard their privacy and security and help them achieve their goals.
The research in this thesis also explores educators’ perceived challenges and interests in classroom integration of generative AI. Key findings include the perceived necessity for both teachers and students to learn about generative AI, the interest across different levels of education and that the age deemed appropriate for students to be introduced to generative AI is often considered to be at the end of primary school or at the start of secondary school.
From the research several design principles are proposed to create an engaging, ethical and human-centred AI learning experience. To facilitate this experience and to spearhead the integration of generative AI into education, a company named “PLAI” is conceptualised. The company aims to implement generative AI in education in a positive way by providing learning material which stimulates engagement, creativity and social learning while safeguarding privacy and security by offering play-based and scaffolded learning in workshops. Future recommendations and plans include the development of a multi-modal generative AI model which can run locally on school servers and is alignable with their curriculum.
Limitations of this study include a lack of direct interaction and research with high school students, a need to explore text based generative AI interactions with students, and a need to assess generative AI’s long-term effects on student well-being.
The study captures the diverse perceptions of elementary school children and teachers regarding generative AI. Key findings highlight children’s concerns and needs with regards to generative AI. Cooperatively formulated design criteria point towards children’s main interests for products or services with generative AI or AI to have a play-based nature, foster creativity and inspiration, safeguard their privacy and security and help them achieve their goals.
The research in this thesis also explores educators’ perceived challenges and interests in classroom integration of generative AI. Key findings include the perceived necessity for both teachers and students to learn about generative AI, the interest across different levels of education and that the age deemed appropriate for students to be introduced to generative AI is often considered to be at the end of primary school or at the start of secondary school.
From the research several design principles are proposed to create an engaging, ethical and human-centred AI learning experience. To facilitate this experience and to spearhead the integration of generative AI into education, a company named “PLAI” is conceptualised. The company aims to implement generative AI in education in a positive way by providing learning material which stimulates engagement, creativity and social learning while safeguarding privacy and security by offering play-based and scaffolded learning in workshops. Future recommendations and plans include the development of a multi-modal generative AI model which can run locally on school servers and is alignable with their curriculum.
Limitations of this study include a lack of direct interaction and research with high school students, a need to explore text based generative AI interactions with students, and a need to assess generative AI’s long-term effects on student well-being.
Background
This master’s thesis focuses on developing a
game that utilizes ChatGPT and the Sciffle Box to guide players in discussing their
understanding of corporate values and connecting more deeply with corporate values
based on their own human values. Values are abstract and challenging to
comprehend. In corporate value training, personal values are often disregarded,
resulting in a reduced impact. This project aims to unpack these abstract human
values, rendering them more accessible for understanding and communication. By
integrating ChatGPT and the Sciffle Box, value communication can become more
playful and immersive. Through the connection to their own values, the design has
the potential to create a greater impact on employees, fostering a deeper connection
with their corporate values.
Research and design
The study employs a five-stage methodology to
investigate the symbiotic potential of the Sciffle Box, human values, and value
communication through gaming. This journey involves a literature review focused
on understanding values and corporate values, along with research on the
Sciffle Box and its contextual use. Three co-design sessions are conducted with
master students at Delft University of Technology, identifying various contexts
using a value based Sciffle Box game. Insights are synthesized into an
opportunity map, followed by iterative design of five prototypes, culminating
in the deployment of the most effective design. The final prototype is built
with Unity on the Sciffle Box, evaluated by seven staff of Ijsfontein. The
final design mapped out five different working dilemmas that are merged in five
narratives presented by non-game players played by ChatGPT. Each dilemma has a
behaviour goal which matches the corporate values. Players are divided into
four teams; each team will have five different human values cards that can only
be used once. By applying those values cards, players need to discuss with their
team on what is their understanding of corporate values. In the end, each team
will present their solution and vote for the best value that can solve the
dilemma and match with corporate values. Non-game players will provide feedback
based on the value that got the most votes.
Evaluation and discussion
The integration of the Sciffle Box and ChatGPT elevates user engagement and comprehension of values. The application of different value theory, demonstrated in the selection and categorization of values, ensures a comprehensive spectrum is represented, fostering nuanced, meaningful discussions among participants. The utilization of ChatGPT showcases AI’s potential in enhancing value-focused game play, generating contextually relevant feedback, and enabling valuable, interactive dialogues. Moreover, the project bridges the gap between personal and corporate values found in conventional corporate training, prompting users to align their individual values with corporate ones, a catalyst for behavioural changes that harmonize personal aspirations with organizational objectives. While limitations do exist, such as the finite range of value cards and the reliance on ChatGPT’s feedback accuracy, the project’s key findings underscore substantial potential for the Sciffle Box as a tool for effective value communication and awareness in a corporate context. Further research can be done to explore how to use AI technology in various ways (text/ audio/video) in corporate training serious games, how to stabilize its performance, and support meaningful conversations. Simultaneously, additional research can discover more value structures that can support value communication in context. ...
Background
Research and design
Evaluation and discussion
This master’s thesis focuses on developing a game that utilizes ChatGPT and the Sciffle Box to guide players in discussing their understanding of corporate values and connecting more deeply with corporate values based on their own human values. Values are abstract and challenging to comprehend. In corporate value training, personal values are often disregarded, resulting in a reduced impact. This project aims to unpack these abstract human values, rendering them more accessible for understanding and communication. By integrating ChatGPT and the Sciffle Box, value communication can become more playful and immersive. Through the connection to their own values, the design has the potential to create a greater impact on employees, fostering a deeper connection with their corporate values.
Research and design
The study employs a five-stage methodology to investigate the symbiotic potential of the Sciffle Box, human values, and value communication through gaming. This journey involves a literature review focused on understanding values and corporate values, along with research on the Sciffle Box and its contextual use. Three co-design sessions are conducted with master students at Delft University of Technology, identifying various contexts using a value based Sciffle Box game. Insights are synthesized into an opportunity map, followed by iterative design of five prototypes, culminating in the deployment of the most effective design. The final prototype is built with Unity on the Sciffle Box, evaluated by seven staff of Ijsfontein. The final design mapped out five different working dilemmas that are merged in five narratives presented by non-game players played by ChatGPT. Each dilemma has a behaviour goal which matches the corporate values. Players are divided into four teams; each team will have five different human values cards that can only be used once. By applying those values cards, players need to discuss with their team on what is their understanding of corporate values. In the end, each team will present their solution and vote for the best value that can solve the dilemma and match with corporate values. Non-game players will provide feedback based on the value that got the most votes.
Evaluation and discussion
The integration of the Sciffle Box and ChatGPT elevates user engagement and comprehension of values. The application of different value theory, demonstrated in the selection and categorization of values, ensures a comprehensive spectrum is represented, fostering nuanced, meaningful discussions among participants. The utilization of ChatGPT showcases AI’s potential in enhancing value-focused game play, generating contextually relevant feedback, and enabling valuable, interactive dialogues. Moreover, the project bridges the gap between personal and corporate values found in conventional corporate training, prompting users to align their individual values with corporate ones, a catalyst for behavioural changes that harmonize personal aspirations with organizational objectives. While limitations do exist, such as the finite range of value cards and the reliance on ChatGPT’s feedback accuracy, the project’s key findings underscore substantial potential for the Sciffle Box as a tool for effective value communication and awareness in a corporate context. Further research can be done to explore how to use AI technology in various ways (text/ audio/video) in corporate training serious games, how to stabilize its performance, and support meaningful conversations. Simultaneously, additional research can discover more value structures that can support value communication in context.
Design for Diverse Body Shapes with AI
Exploring AIGC Applications for DINED Mannequin Generation
What
In this project, a workflow method integrating generative AI has been developed. This versatile approach can be applied to a variety of body shape-related design tasks. For instance, designers can leverage this method effectively to generate design inspirations, enabling them to establish detailed design themes and goals more efficiently. Building upon this workflow, a new tool named DINED AI has been specifically designed to offer a more user-friendly and guided approach to accomplish these tasks. Although DINED AI is currently in the form of functional prototypes, both the user interface and back-end technology have been individually crafted to provide a realistic demonstration of its capabilities.
Why
With the rapid advancements in large language models, integrating AI technologies has emerged as the prevailing trend in product development. The aim of this project is to explore the potential of creating an AI-powered platform for DINED. DINED is renowned for its specialized expertise in anthropometry, which allows for the generation of realistic data models to aid in the design of ergonomic products. Leveraging this distinctive capability of DINED, and under the guidance of the mentors, this project focuses on the design of clothing, a widely recognized product that is intimately connected to the human body shape's form. Additionally, a streamlined workflow method will be developed to facilitate the design process.
How
The project is centered around leveraging technological advancements by implementing Stable Diffusion, LoRA, and ControlNet as crucial components. Stable Diffusion is a well-established text-to-image algorithm, while the LoRA model effectively refines Stable Diffusion using a limited number of images in the training dataset. On the other hand, ControlNet possesses the unique capability to extract specific control elements from the 3D mannequin generated by DINED, thereby enabling enhanced control over the resulting images.
In order to ensure a seamless user experience, the project integrates these aforementioned technologies into a cohesive and streamlined workflow. Each step of the process is accompanied by comprehensive guidance, empowering users to effortlessly navigate the system and fostering a coherent user experience. Furthermore, the project underwent evaluation through expert interviews, further validating its effectiveness and potential impact. ...
In this project, a workflow method integrating generative AI has been developed. This versatile approach can be applied to a variety of body shape-related design tasks. For instance, designers can leverage this method effectively to generate design inspirations, enabling them to establish detailed design themes and goals more efficiently. Building upon this workflow, a new tool named DINED AI has been specifically designed to offer a more user-friendly and guided approach to accomplish these tasks. Although DINED AI is currently in the form of functional prototypes, both the user interface and back-end technology have been individually crafted to provide a realistic demonstration of its capabilities.
Why
With the rapid advancements in large language models, integrating AI technologies has emerged as the prevailing trend in product development. The aim of this project is to explore the potential of creating an AI-powered platform for DINED. DINED is renowned for its specialized expertise in anthropometry, which allows for the generation of realistic data models to aid in the design of ergonomic products. Leveraging this distinctive capability of DINED, and under the guidance of the mentors, this project focuses on the design of clothing, a widely recognized product that is intimately connected to the human body shape's form. Additionally, a streamlined workflow method will be developed to facilitate the design process.
How
The project is centered around leveraging technological advancements by implementing Stable Diffusion, LoRA, and ControlNet as crucial components. Stable Diffusion is a well-established text-to-image algorithm, while the LoRA model effectively refines Stable Diffusion using a limited number of images in the training dataset. On the other hand, ControlNet possesses the unique capability to extract specific control elements from the 3D mannequin generated by DINED, thereby enabling enhanced control over the resulting images.
In order to ensure a seamless user experience, the project integrates these aforementioned technologies into a cohesive and streamlined workflow. Each step of the process is accompanied by comprehensive guidance, empowering users to effortlessly navigate the system and fostering a coherent user experience. Furthermore, the project underwent evaluation through expert interviews, further validating its effectiveness and potential impact. ...
What
In this project, a workflow method integrating generative AI has been developed. This versatile approach can be applied to a variety of body shape-related design tasks. For instance, designers can leverage this method effectively to generate design inspirations, enabling them to establish detailed design themes and goals more efficiently. Building upon this workflow, a new tool named DINED AI has been specifically designed to offer a more user-friendly and guided approach to accomplish these tasks. Although DINED AI is currently in the form of functional prototypes, both the user interface and back-end technology have been individually crafted to provide a realistic demonstration of its capabilities.
Why
With the rapid advancements in large language models, integrating AI technologies has emerged as the prevailing trend in product development. The aim of this project is to explore the potential of creating an AI-powered platform for DINED. DINED is renowned for its specialized expertise in anthropometry, which allows for the generation of realistic data models to aid in the design of ergonomic products. Leveraging this distinctive capability of DINED, and under the guidance of the mentors, this project focuses on the design of clothing, a widely recognized product that is intimately connected to the human body shape's form. Additionally, a streamlined workflow method will be developed to facilitate the design process.
How
The project is centered around leveraging technological advancements by implementing Stable Diffusion, LoRA, and ControlNet as crucial components. Stable Diffusion is a well-established text-to-image algorithm, while the LoRA model effectively refines Stable Diffusion using a limited number of images in the training dataset. On the other hand, ControlNet possesses the unique capability to extract specific control elements from the 3D mannequin generated by DINED, thereby enabling enhanced control over the resulting images.
In order to ensure a seamless user experience, the project integrates these aforementioned technologies into a cohesive and streamlined workflow. Each step of the process is accompanied by comprehensive guidance, empowering users to effortlessly navigate the system and fostering a coherent user experience. Furthermore, the project underwent evaluation through expert interviews, further validating its effectiveness and potential impact.
In this project, a workflow method integrating generative AI has been developed. This versatile approach can be applied to a variety of body shape-related design tasks. For instance, designers can leverage this method effectively to generate design inspirations, enabling them to establish detailed design themes and goals more efficiently. Building upon this workflow, a new tool named DINED AI has been specifically designed to offer a more user-friendly and guided approach to accomplish these tasks. Although DINED AI is currently in the form of functional prototypes, both the user interface and back-end technology have been individually crafted to provide a realistic demonstration of its capabilities.
Why
With the rapid advancements in large language models, integrating AI technologies has emerged as the prevailing trend in product development. The aim of this project is to explore the potential of creating an AI-powered platform for DINED. DINED is renowned for its specialized expertise in anthropometry, which allows for the generation of realistic data models to aid in the design of ergonomic products. Leveraging this distinctive capability of DINED, and under the guidance of the mentors, this project focuses on the design of clothing, a widely recognized product that is intimately connected to the human body shape's form. Additionally, a streamlined workflow method will be developed to facilitate the design process.
How
The project is centered around leveraging technological advancements by implementing Stable Diffusion, LoRA, and ControlNet as crucial components. Stable Diffusion is a well-established text-to-image algorithm, while the LoRA model effectively refines Stable Diffusion using a limited number of images in the training dataset. On the other hand, ControlNet possesses the unique capability to extract specific control elements from the 3D mannequin generated by DINED, thereby enabling enhanced control over the resulting images.
In order to ensure a seamless user experience, the project integrates these aforementioned technologies into a cohesive and streamlined workflow. Each step of the process is accompanied by comprehensive guidance, empowering users to effortlessly navigate the system and fostering a coherent user experience. Furthermore, the project underwent evaluation through expert interviews, further validating its effectiveness and potential impact.
New generative AI systems can make beautiful pictures, but can they help produce functional products? This graduation project investigated the potential for generative AI to aid the design of functional, 3D-printed lamps.
Two studies were conducted to assess the degree to which AI can measurably improve design outputs. An AI-based software system was developed to help designers explore the possibilities of lamp design. Output images from this system were used as inputs to generate 3D models, in an example of human-AI collaboration, then 3D models were then printed using four different 3D printing technologies.
Altogether, this work showcases the potential for AI to translate human imagination into physical products. Despite meaningful challenges in the process of 3D printing and 2D to 3D conversions, AI advancements signal a promising future for the design industry—and a need for new approaches to design education. ...
Two studies were conducted to assess the degree to which AI can measurably improve design outputs. An AI-based software system was developed to help designers explore the possibilities of lamp design. Output images from this system were used as inputs to generate 3D models, in an example of human-AI collaboration, then 3D models were then printed using four different 3D printing technologies.
Altogether, this work showcases the potential for AI to translate human imagination into physical products. Despite meaningful challenges in the process of 3D printing and 2D to 3D conversions, AI advancements signal a promising future for the design industry—and a need for new approaches to design education. ...
New generative AI systems can make beautiful pictures, but can they help produce functional products? This graduation project investigated the potential for generative AI to aid the design of functional, 3D-printed lamps.
Two studies were conducted to assess the degree to which AI can measurably improve design outputs. An AI-based software system was developed to help designers explore the possibilities of lamp design. Output images from this system were used as inputs to generate 3D models, in an example of human-AI collaboration, then 3D models were then printed using four different 3D printing technologies.
Altogether, this work showcases the potential for AI to translate human imagination into physical products. Despite meaningful challenges in the process of 3D printing and 2D to 3D conversions, AI advancements signal a promising future for the design industry—and a need for new approaches to design education.
Two studies were conducted to assess the degree to which AI can measurably improve design outputs. An AI-based software system was developed to help designers explore the possibilities of lamp design. Output images from this system were used as inputs to generate 3D models, in an example of human-AI collaboration, then 3D models were then printed using four different 3D printing technologies.
Altogether, this work showcases the potential for AI to translate human imagination into physical products. Despite meaningful challenges in the process of 3D printing and 2D to 3D conversions, AI advancements signal a promising future for the design industry—and a need for new approaches to design education.
The market of quantum computing is growing rapidly due to the rising demand of high-performance computing. By combining quantum theory and computer science, quantum computing shows great promise for future application in fields like algorithms, cryptography, machine learning and quantum simulation. Current breakthroughs in quantum technology and rising investments in the quantum computing market indicate a high demand of quantum physicists over the coming decade.
Looking at projected future investments in the quantum computing market, the current state of the quantum technology education infrastructure and it’s foreseeable shortcoming in providing a sufficiently strong workforce to answer the industry’s expansion, it can be expected that quantum education will be pressurized to conform to big change to keep up with the industry.
The project aims to provide a solution that will help open ways to achieve maintaining the balance of education and industry that is needed within the quantum technology sector if the industry keeps expanding as it is currently. It hopes to deliver where currently is missing: an engaging way to interest our current youth in quantum technology before they reach university, in a way that motivates them to become one of the future quantum physicists that the industry will drastically need.
Currently basic aspects of quantum physics are difficult to convey to a younger generation in a compelling and engaging manner. This project aimed to introduce basic aspects of quantum physics through a science museum exhibit using cymatics, the study of visual wave phenomena. The wave-like characteristic found in both quantum physics and cymatics was the primary factor to investigate how cymatics could be used to introduce basic aspects of quantum physics to young future scientists through a science museum exhibit design. The modes of vibration in cymatics are called eigenmodes.
The final design is an exhibit design that uses these eigenmodes to simulate how quantum physicists work with their delicate and sensitive quantum systems. In the design the eigenmodes resemble the sensitivity of the qubits that these physicists work with.
...
Looking at projected future investments in the quantum computing market, the current state of the quantum technology education infrastructure and it’s foreseeable shortcoming in providing a sufficiently strong workforce to answer the industry’s expansion, it can be expected that quantum education will be pressurized to conform to big change to keep up with the industry.
The project aims to provide a solution that will help open ways to achieve maintaining the balance of education and industry that is needed within the quantum technology sector if the industry keeps expanding as it is currently. It hopes to deliver where currently is missing: an engaging way to interest our current youth in quantum technology before they reach university, in a way that motivates them to become one of the future quantum physicists that the industry will drastically need.
Currently basic aspects of quantum physics are difficult to convey to a younger generation in a compelling and engaging manner. This project aimed to introduce basic aspects of quantum physics through a science museum exhibit using cymatics, the study of visual wave phenomena. The wave-like characteristic found in both quantum physics and cymatics was the primary factor to investigate how cymatics could be used to introduce basic aspects of quantum physics to young future scientists through a science museum exhibit design. The modes of vibration in cymatics are called eigenmodes.
The final design is an exhibit design that uses these eigenmodes to simulate how quantum physicists work with their delicate and sensitive quantum systems. In the design the eigenmodes resemble the sensitivity of the qubits that these physicists work with.
...
The market of quantum computing is growing rapidly due to the rising demand of high-performance computing. By combining quantum theory and computer science, quantum computing shows great promise for future application in fields like algorithms, cryptography, machine learning and quantum simulation. Current breakthroughs in quantum technology and rising investments in the quantum computing market indicate a high demand of quantum physicists over the coming decade.
Looking at projected future investments in the quantum computing market, the current state of the quantum technology education infrastructure and it’s foreseeable shortcoming in providing a sufficiently strong workforce to answer the industry’s expansion, it can be expected that quantum education will be pressurized to conform to big change to keep up with the industry.
The project aims to provide a solution that will help open ways to achieve maintaining the balance of education and industry that is needed within the quantum technology sector if the industry keeps expanding as it is currently. It hopes to deliver where currently is missing: an engaging way to interest our current youth in quantum technology before they reach university, in a way that motivates them to become one of the future quantum physicists that the industry will drastically need.
Currently basic aspects of quantum physics are difficult to convey to a younger generation in a compelling and engaging manner. This project aimed to introduce basic aspects of quantum physics through a science museum exhibit using cymatics, the study of visual wave phenomena. The wave-like characteristic found in both quantum physics and cymatics was the primary factor to investigate how cymatics could be used to introduce basic aspects of quantum physics to young future scientists through a science museum exhibit design. The modes of vibration in cymatics are called eigenmodes.
The final design is an exhibit design that uses these eigenmodes to simulate how quantum physicists work with their delicate and sensitive quantum systems. In the design the eigenmodes resemble the sensitivity of the qubits that these physicists work with.
Looking at projected future investments in the quantum computing market, the current state of the quantum technology education infrastructure and it’s foreseeable shortcoming in providing a sufficiently strong workforce to answer the industry’s expansion, it can be expected that quantum education will be pressurized to conform to big change to keep up with the industry.
The project aims to provide a solution that will help open ways to achieve maintaining the balance of education and industry that is needed within the quantum technology sector if the industry keeps expanding as it is currently. It hopes to deliver where currently is missing: an engaging way to interest our current youth in quantum technology before they reach university, in a way that motivates them to become one of the future quantum physicists that the industry will drastically need.
Currently basic aspects of quantum physics are difficult to convey to a younger generation in a compelling and engaging manner. This project aimed to introduce basic aspects of quantum physics through a science museum exhibit using cymatics, the study of visual wave phenomena. The wave-like characteristic found in both quantum physics and cymatics was the primary factor to investigate how cymatics could be used to introduce basic aspects of quantum physics to young future scientists through a science museum exhibit design. The modes of vibration in cymatics are called eigenmodes.
The final design is an exhibit design that uses these eigenmodes to simulate how quantum physicists work with their delicate and sensitive quantum systems. In the design the eigenmodes resemble the sensitivity of the qubits that these physicists work with.
The problem we want to address
AI(Artificial intelligence) is diving into people’s lives as its algorithm continues to iterate. However, the algorithmic and quantitative systems do not seem to access people’s experiences, which always include emotional and qualitative factors. How can an AI system understand people’s feelings? How can an AI system like GANs optimize for meaningful human values, like beauty?
Our approach
In this project, I mainly build two approaches for designers to translate the human experience into GANs. Starting from the exploration of GANs, I investigate their potential chances to “understand” human experience. Then, the qualitative research
on human experience reveals the aesthetic factors influencing people’s feelings
about GANs‘ output. Through the interviews, some pain points from AI product
designers are concluded. Combining the exploration results in those three scopes, I
hypothesize that retraining with the highly-rated images and building new computational models to iterate the factors from the human evaluation can help designers inform the AI system about the human experience. I cooperated with CSE students and proposed two approaches- CURATION APPROACH and ALGORITHMIC
AESTHETIC APPROACH- based on the above assumptions. The build process demonstrates the achievability of the assumptions. In the evaluation part, the retrained images in both methods are rated higher than the original ones. Both approaches are proved to be practical and feasible.
Our result
For the curation approach, all produced models with highly-rated images outperformed the original model. The original model’s score is 0.2. When replacing 500,1000,2985 highly-rated images into the input dataset, their scores increased to 0.245, 0.269, and 0.255. For the algorithmic aesthetic approach, we select three categories of images(coastline, forest/desert, arctic). In each category of images, three factors are selected to improve. Among the nine groups whose corresponding factors are iterated by algorithms, eight groups’ new images are rated higher by people than the original images. The curation and algorithmic aesthetic approaches are verified to help inform AI systems of human experience.
Potential impact
Following the curation and algorithmic approaches, designers can successfully inform GANs about people’s aesthetic evaluation of their images. Using the human experience to improve AI systems is a starting point. It also proves the importance of the human experience for AI and provides a template for designers in all AI fields to inform their AI systems of human experience. ...
AI(Artificial intelligence) is diving into people’s lives as its algorithm continues to iterate. However, the algorithmic and quantitative systems do not seem to access people’s experiences, which always include emotional and qualitative factors. How can an AI system understand people’s feelings? How can an AI system like GANs optimize for meaningful human values, like beauty?
Our approach
In this project, I mainly build two approaches for designers to translate the human experience into GANs. Starting from the exploration of GANs, I investigate their potential chances to “understand” human experience. Then, the qualitative research
on human experience reveals the aesthetic factors influencing people’s feelings
about GANs‘ output. Through the interviews, some pain points from AI product
designers are concluded. Combining the exploration results in those three scopes, I
hypothesize that retraining with the highly-rated images and building new computational models to iterate the factors from the human evaluation can help designers inform the AI system about the human experience. I cooperated with CSE students and proposed two approaches- CURATION APPROACH and ALGORITHMIC
AESTHETIC APPROACH- based on the above assumptions. The build process demonstrates the achievability of the assumptions. In the evaluation part, the retrained images in both methods are rated higher than the original ones. Both approaches are proved to be practical and feasible.
Our result
For the curation approach, all produced models with highly-rated images outperformed the original model. The original model’s score is 0.2. When replacing 500,1000,2985 highly-rated images into the input dataset, their scores increased to 0.245, 0.269, and 0.255. For the algorithmic aesthetic approach, we select three categories of images(coastline, forest/desert, arctic). In each category of images, three factors are selected to improve. Among the nine groups whose corresponding factors are iterated by algorithms, eight groups’ new images are rated higher by people than the original images. The curation and algorithmic aesthetic approaches are verified to help inform AI systems of human experience.
Potential impact
Following the curation and algorithmic approaches, designers can successfully inform GANs about people’s aesthetic evaluation of their images. Using the human experience to improve AI systems is a starting point. It also proves the importance of the human experience for AI and provides a template for designers in all AI fields to inform their AI systems of human experience. ...
The problem we want to address
AI(Artificial intelligence) is diving into people’s lives as its algorithm continues to iterate. However, the algorithmic and quantitative systems do not seem to access people’s experiences, which always include emotional and qualitative factors. How can an AI system understand people’s feelings? How can an AI system like GANs optimize for meaningful human values, like beauty?
Our approach
In this project, I mainly build two approaches for designers to translate the human experience into GANs. Starting from the exploration of GANs, I investigate their potential chances to “understand” human experience. Then, the qualitative research
on human experience reveals the aesthetic factors influencing people’s feelings
about GANs‘ output. Through the interviews, some pain points from AI product
designers are concluded. Combining the exploration results in those three scopes, I
hypothesize that retraining with the highly-rated images and building new computational models to iterate the factors from the human evaluation can help designers inform the AI system about the human experience. I cooperated with CSE students and proposed two approaches- CURATION APPROACH and ALGORITHMIC
AESTHETIC APPROACH- based on the above assumptions. The build process demonstrates the achievability of the assumptions. In the evaluation part, the retrained images in both methods are rated higher than the original ones. Both approaches are proved to be practical and feasible.
Our result
For the curation approach, all produced models with highly-rated images outperformed the original model. The original model’s score is 0.2. When replacing 500,1000,2985 highly-rated images into the input dataset, their scores increased to 0.245, 0.269, and 0.255. For the algorithmic aesthetic approach, we select three categories of images(coastline, forest/desert, arctic). In each category of images, three factors are selected to improve. Among the nine groups whose corresponding factors are iterated by algorithms, eight groups’ new images are rated higher by people than the original images. The curation and algorithmic aesthetic approaches are verified to help inform AI systems of human experience.
Potential impact
Following the curation and algorithmic approaches, designers can successfully inform GANs about people’s aesthetic evaluation of their images. Using the human experience to improve AI systems is a starting point. It also proves the importance of the human experience for AI and provides a template for designers in all AI fields to inform their AI systems of human experience.
AI(Artificial intelligence) is diving into people’s lives as its algorithm continues to iterate. However, the algorithmic and quantitative systems do not seem to access people’s experiences, which always include emotional and qualitative factors. How can an AI system understand people’s feelings? How can an AI system like GANs optimize for meaningful human values, like beauty?
Our approach
In this project, I mainly build two approaches for designers to translate the human experience into GANs. Starting from the exploration of GANs, I investigate their potential chances to “understand” human experience. Then, the qualitative research
on human experience reveals the aesthetic factors influencing people’s feelings
about GANs‘ output. Through the interviews, some pain points from AI product
designers are concluded. Combining the exploration results in those three scopes, I
hypothesize that retraining with the highly-rated images and building new computational models to iterate the factors from the human evaluation can help designers inform the AI system about the human experience. I cooperated with CSE students and proposed two approaches- CURATION APPROACH and ALGORITHMIC
AESTHETIC APPROACH- based on the above assumptions. The build process demonstrates the achievability of the assumptions. In the evaluation part, the retrained images in both methods are rated higher than the original ones. Both approaches are proved to be practical and feasible.
Our result
For the curation approach, all produced models with highly-rated images outperformed the original model. The original model’s score is 0.2. When replacing 500,1000,2985 highly-rated images into the input dataset, their scores increased to 0.245, 0.269, and 0.255. For the algorithmic aesthetic approach, we select three categories of images(coastline, forest/desert, arctic). In each category of images, three factors are selected to improve. Among the nine groups whose corresponding factors are iterated by algorithms, eight groups’ new images are rated higher by people than the original images. The curation and algorithmic aesthetic approaches are verified to help inform AI systems of human experience.
Potential impact
Following the curation and algorithmic approaches, designers can successfully inform GANs about people’s aesthetic evaluation of their images. Using the human experience to improve AI systems is a starting point. It also proves the importance of the human experience for AI and provides a template for designers in all AI fields to inform their AI systems of human experience.
Bachelor thesis
(2022)
-
T.H. Nguyen, J.D. Lomas, U.K. Gadiraju, W.L.A. van der Maden, G.M. Allen, D.H.J. Epema
Despite the fact that climate change is becoming increasingly dangerous and prevalent, there is still a lack of public engagement. This can be explained by the fact that the media portrays climate change as an abstract concept. The message can be more effectively communicated through visual art because it is more likely to invoke emotional responses in individuals. By including human perception and rating data, the generative adversarial neural network (GAN) produces better image output. Therefore, this paper explores methods for using the human perception of beauty in order to improve StyleGAN2 outputs. In GANAesthetic, UI sliders allow users to explore satellite images interactively, that is, visually appealing satellite images generated from StyleGAN2. The GANAesthetic was determined to be the most appropriate methodology for the study. The choice of GANAesthetic over other approaches will be explained in this paper, as well as its implementation. The paper will also describe an experiment to discover aesthetic latent dimensions.
...
Despite the fact that climate change is becoming increasingly dangerous and prevalent, there is still a lack of public engagement. This can be explained by the fact that the media portrays climate change as an abstract concept. The message can be more effectively communicated through visual art because it is more likely to invoke emotional responses in individuals. By including human perception and rating data, the generative adversarial neural network (GAN) produces better image output. Therefore, this paper explores methods for using the human perception of beauty in order to improve StyleGAN2 outputs. In GANAesthetic, UI sliders allow users to explore satellite images interactively, that is, visually appealing satellite images generated from StyleGAN2. The GANAesthetic was determined to be the most appropriate methodology for the study. The choice of GANAesthetic over other approaches will be explained in this paper, as well as its implementation. The paper will also describe an experiment to discover aesthetic latent dimensions.
Bachelor thesis
(2022)
-
M. Rahman, J.D. Lomas, U.K. Gadiraju, W.L.A. van der Maden, G.M. Allen, D.H.J. Epema
Generative Adversarial Networks (GANs) can create artwork images and we need effective ways of rating their aesthetic values. This could help us determine the most aesthetic artwork images (and identify the GANs that created them) and train GANs to produce more aesthetic artwork images in the future. In this research, we analyzed the effectiveness of using two different survey formats (binary-choice and four-choice) for displaying GAN-produced artwork images to crowd- sourced workers and gathering their ratings. The artwork images were of different landscapes like the desert, arctic, coastal regions, etc. Additionally, we investigated how the choice of showing different images together (image groupings) per question affects the final rating results. Results demonstrate that the four-choice format is superior to the binary-choice format in producing more consistent, reliable, and accurate results. The effects of the different image groupings were insignif- icant for the results of the four-choice format. In contrast, different image groupings displayed statistically significant changes in the results for the binary-choice format. However, it was found that crowdsourced workers preferred the binary-choice format more as they found it to be less strenuous and more effective in allowing them to express their rating choices.
...
Generative Adversarial Networks (GANs) can create artwork images and we need effective ways of rating their aesthetic values. This could help us determine the most aesthetic artwork images (and identify the GANs that created them) and train GANs to produce more aesthetic artwork images in the future. In this research, we analyzed the effectiveness of using two different survey formats (binary-choice and four-choice) for displaying GAN-produced artwork images to crowd- sourced workers and gathering their ratings. The artwork images were of different landscapes like the desert, arctic, coastal regions, etc. Additionally, we investigated how the choice of showing different images together (image groupings) per question affects the final rating results. Results demonstrate that the four-choice format is superior to the binary-choice format in producing more consistent, reliable, and accurate results. The effects of the different image groupings were insignif- icant for the results of the four-choice format. In contrast, different image groupings displayed statistically significant changes in the results for the binary-choice format. However, it was found that crowdsourced workers preferred the binary-choice format more as they found it to be less strenuous and more effective in allowing them to express their rating choices.
Iterative training with human rated images to improve GAN generated image aesthetics
Effects of dataset size and training length
Bachelor thesis
(2022)
-
B.I. Çelebi, W.L.A. van der Maden, J.D. Lomas, G.M. Allen, U.K. Gadiraju, D.H.J. Epema
Generative Adversarial Networks (GANs) brought rapid developments in generating synthetic images by mimicking structures in the training data. With the list of application of GANs growing drastically, it has lately become an exciting technology to explore for designers to communicate their ideas and arts through technology and create engaging experiences for humans. Nevertheless, translating human experiences to artificial intelligence and creating visually pleasant imagery is a challenging task due to complex semantics of human perception. To address this issue, we introduce an iterative training approach in which the generated images are curated by humans and the most pleasing ones are fed back into the network to retrain. Additionally, we do a factorial analysis to investigate how the aesthetic quality and the diversity are affected by the size of training data and training length. In experiments, we validate that this method can significantly improve the aesthetic quality of generated images regardless of the dataset size and training length, however the use of smaller datasets comes with a cost of reduction in the image diversity and novelty in the output images. The aesthetic bias towards certain contexts can also deteriorate the diversity and affect the model evaluations. On the other hand, no significant relationship has been found regarding the training length, however this could possibly be due to instabilities that happen during the model convergence progress.
...
Generative Adversarial Networks (GANs) brought rapid developments in generating synthetic images by mimicking structures in the training data. With the list of application of GANs growing drastically, it has lately become an exciting technology to explore for designers to communicate their ideas and arts through technology and create engaging experiences for humans. Nevertheless, translating human experiences to artificial intelligence and creating visually pleasant imagery is a challenging task due to complex semantics of human perception. To address this issue, we introduce an iterative training approach in which the generated images are curated by humans and the most pleasing ones are fed back into the network to retrain. Additionally, we do a factorial analysis to investigate how the aesthetic quality and the diversity are affected by the size of training data and training length. In experiments, we validate that this method can significantly improve the aesthetic quality of generated images regardless of the dataset size and training length, however the use of smaller datasets comes with a cost of reduction in the image diversity and novelty in the output images. The aesthetic bias towards certain contexts can also deteriorate the diversity and affect the model evaluations. On the other hand, no significant relationship has been found regarding the training length, however this could possibly be due to instabilities that happen during the model convergence progress.
Beauty in the Eye of Machine
Using Automated Measures of Aesthetic Beauty to Improve GAN Output of Satellite Images
Bachelor thesis
(2022)
-
J.M. Catlett, J.D. Lomas, W.L.A. van der Maden, U.K. Gadiraju, G.M. Allen, D.H.J. Epema
This paper aims to evaluate which automated measures of aesthetic beauty are the best predictors for human ratings of aesthetics and proposes that typicality and novelty may increase the correlation between the two. To study the correlation between these metrics, a literature study was performed to find a select amount of potentially good predictors, a pipeline was created to extract these values from each image within our dataset, a survey was conducted to vote for which images were considered most aesthetic, and finally regression analysis was performed to see which metrics offered highest correlation with the human rating data. From this we could see there were indeed a number of automated metrics that consistently scored high as predictors for the human aesthetic ratings and there was a slight improvement in the fit of the prediction model upon including novelty as a feature. However, at this moment, the improvement is not significant to conclude these features are better at predicting human ratings.
...
This paper aims to evaluate which automated measures of aesthetic beauty are the best predictors for human ratings of aesthetics and proposes that typicality and novelty may increase the correlation between the two. To study the correlation between these metrics, a literature study was performed to find a select amount of potentially good predictors, a pipeline was created to extract these values from each image within our dataset, a survey was conducted to vote for which images were considered most aesthetic, and finally regression analysis was performed to see which metrics offered highest correlation with the human rating data. From this we could see there were indeed a number of automated metrics that consistently scored high as predictors for the human aesthetic ratings and there was a slight improvement in the fit of the prediction model upon including novelty as a feature. However, at this moment, the improvement is not significant to conclude these features are better at predicting human ratings.
Bachelor thesis
(2022)
-
A. Achilleos, W.L.A. van der Maden, J.D. Lomas, U.K. Gadiraju, G.M. Allen, Z. Erkin
With the COVID-19 pandemic testing humanity worldwide in unforeseen ways, wellbeing assessment has stepped to the foreground of individual health status. Conversational User Interfaces (CUIs) prove promising as an assessment tool, but lacking the means to retain users engaged during the process. This research aims to explore a solution introducing a graphical modality to a CUI, that improves user experience during wellbeing assessment. A psychological projective testing technique was researched, the H-T-P test, which exposes one's personality, mood and feelings through drawing. A system was designed that utilizes a simplified version of the test and image classification to provide a new means of wellbeing assessment. The system's performance and usability show that such an approach is indeed feasible. This paper describes the design process of the system, the resulting prototype and possible improvements for future implementations.
...
With the COVID-19 pandemic testing humanity worldwide in unforeseen ways, wellbeing assessment has stepped to the foreground of individual health status. Conversational User Interfaces (CUIs) prove promising as an assessment tool, but lacking the means to retain users engaged during the process. This research aims to explore a solution introducing a graphical modality to a CUI, that improves user experience during wellbeing assessment. A psychological projective testing technique was researched, the H-T-P test, which exposes one's personality, mood and feelings through drawing. A system was designed that utilizes a simplified version of the test and image classification to provide a new means of wellbeing assessment. The system's performance and usability show that such an approach is indeed feasible. This paper describes the design process of the system, the resulting prototype and possible improvements for future implementations.