NextBillion - A WDI Publication

AI tools such as ChatGPT, Claude, and Gemini are becoming part of the professional landscape for evaluators and impact measurement practitioners. Much of the conversation about these tools has focused on efficiency and accuracy — whether AI can speed up data collection, automate reporting, or produce reliable analyses. These are important topics, but evaluators do not simply collect, analyze, and report information; their work requires thinking through ambiguity, talking through competing interpretations, and arriving at judgments that are unavoidably subjective. Questions of efficiency and accuracy, then, give us only a partial picture of AI’s implications for evaluation practice. We also need to ask what happens to evaluative reasoning itself when practitioners think through judgment calls with AI.

This webinar will present findings from a qualitative study of 13 impact investing evaluation professionals who incorporated ChatGPT into their work. The research examines what AI can and cannot bring to evaluative dialogue, finding that the very qualities that make AI useful for certain evaluative tasks limit its value for others.

Attendees will come away with a set of practical use cases for incorporating AI dialogue into evaluative work, along with a clearer sense of where AI-mediated reasoning adds value and where human dialogue remains irreplaceable for now. Questions and discussion of personal experiences will be welcomed during the session.