ZW
Z. Wu
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>
1 records found
1
This project investigates how to design a task-level AI adoption decision-support framework that helps UX researchers at Bynder adopt AI responsibly. It responds to a context in which AI adoption is organic and practitioner-led and aims at facilitating the development of standardized governance on AI adoption. The research phase combined a rapid literature scan that identified candidate AI adoption decision constructs, exploratory interviews that surfaced current practice inductively, and construct validation sessions that tested those constructs against UX researchers' own reasoning to confirm or reject those candidate constructs.
These insights were translated into three main interconnecting outcomes: 1. A conceptual reframing of the AI adoption decision from a binary AI use feasibility question into two distinct decisions: whether AI use is feasible, and how human verification of AI output is structured for a given task. 2. A three-layer decision framework operationalizes this through a gating layer, an AI output verification layer that maps task type to suitable verification approaches, and an intensified verification layer that surfaces risk conditions. 3. A prototyped tool, which is grounded in decision support system principles, that translate the framework logic into a tangible tool while preserving researcher judgement rather than automating it.
A formative evaluation engaged UX researchers in think-aloud walkthrough sessions against clarity, usability, and conceptual completeness. Participants found the tool usable, understandable, and coherent and its surfaced factors aligned with their own practice. They also highlighted refinements that were then incorporated into the refined prototype. ...
These insights were translated into three main interconnecting outcomes: 1. A conceptual reframing of the AI adoption decision from a binary AI use feasibility question into two distinct decisions: whether AI use is feasible, and how human verification of AI output is structured for a given task. 2. A three-layer decision framework operationalizes this through a gating layer, an AI output verification layer that maps task type to suitable verification approaches, and an intensified verification layer that surfaces risk conditions. 3. A prototyped tool, which is grounded in decision support system principles, that translate the framework logic into a tangible tool while preserving researcher judgement rather than automating it.
A formative evaluation engaged UX researchers in think-aloud walkthrough sessions against clarity, usability, and conceptual completeness. Participants found the tool usable, understandable, and coherent and its surfaced factors aligned with their own practice. They also highlighted refinements that were then incorporated into the refined prototype. ...
This project investigates how to design a task-level AI adoption decision-support framework that helps UX researchers at Bynder adopt AI responsibly. It responds to a context in which AI adoption is organic and practitioner-led and aims at facilitating the development of standardized governance on AI adoption. The research phase combined a rapid literature scan that identified candidate AI adoption decision constructs, exploratory interviews that surfaced current practice inductively, and construct validation sessions that tested those constructs against UX researchers' own reasoning to confirm or reject those candidate constructs.
These insights were translated into three main interconnecting outcomes: 1. A conceptual reframing of the AI adoption decision from a binary AI use feasibility question into two distinct decisions: whether AI use is feasible, and how human verification of AI output is structured for a given task. 2. A three-layer decision framework operationalizes this through a gating layer, an AI output verification layer that maps task type to suitable verification approaches, and an intensified verification layer that surfaces risk conditions. 3. A prototyped tool, which is grounded in decision support system principles, that translate the framework logic into a tangible tool while preserving researcher judgement rather than automating it.
A formative evaluation engaged UX researchers in think-aloud walkthrough sessions against clarity, usability, and conceptual completeness. Participants found the tool usable, understandable, and coherent and its surfaced factors aligned with their own practice. They also highlighted refinements that were then incorporated into the refined prototype.
These insights were translated into three main interconnecting outcomes: 1. A conceptual reframing of the AI adoption decision from a binary AI use feasibility question into two distinct decisions: whether AI use is feasible, and how human verification of AI output is structured for a given task. 2. A three-layer decision framework operationalizes this through a gating layer, an AI output verification layer that maps task type to suitable verification approaches, and an intensified verification layer that surfaces risk conditions. 3. A prototyped tool, which is grounded in decision support system principles, that translate the framework logic into a tangible tool while preserving researcher judgement rather than automating it.
A formative evaluation engaged UX researchers in think-aloud walkthrough sessions against clarity, usability, and conceptual completeness. Participants found the tool usable, understandable, and coherent and its surfaced factors aligned with their own practice. They also highlighted refinements that were then incorporated into the refined prototype.