Webinar tomorrow: the part of evaluation AI still can't touch
Both our co-founders join the panel to ask what stays human when AI handles the technical work. 3 June, 2 PM IST.
What AI Cannot Replace in Evaluation: The Role of Human Judgement
gLOCAL Evaluation Week 2026 · Session 2 · Tomorrow, 3 June · 2:00–3:30 PM IST
We’ve found ourselves returning to one question lately, working inside systems where evidence shapes programmes, funding, and people’s lives:
What happens when tools built to support human thinking slowly begin to shape it instead?
AI now summarises interviews, structures reports, surfaces themes, and processes large volumes of data in seconds. The efficiency is real. But evaluation rests on context, interpretation, and judgment — the parts that don’t compress into a prompt.
So where’s the line between automation and thinking? At what point does faster output start standing in for deeper understanding?
That’s the conversation in our second gLOCAL Evaluation Week session.
The questions on the table:
What in evaluation fundamentally requires human judgment — and what gets lost when the work is over-automated?
Where do responsibility, credibility, and bias sit when an AI tool drafts the first version?
How is the evaluator’s role changing as more of the technical work shifts to machines?
A preview of what each speaker will touch:
Payal Mulchandani — on field workflows: why an AI-built survey form can look clean on paper but break on day one, and why weak thinking upfront means “AI just scales the wrong thing faster.”
Aditi Chatterjee — on whose voices go missing when AI sets the research agenda, since people with lived experience are largely absent from the data these models are trained on.
Rahul Shah — on the organisational lens: AI can summarise feedback and compare responses, but reading power dynamics and creating space for people to open up is what makes a diagnosis usable.
Sharon Weir — on the line evaluators shouldn’t cross: AI can support the analysis, but it cannot be the analyst.
Who this is for: MEL practitioners and evaluators, researchers and consultants, NGO and CSR teams, programme and OD specialists, and anyone weighing how much of evaluation should be handed to a machine.
This is session two of three in our series, Evaluation, Evidence, and Trust in the Age of AI. You can see details for all three webinars here: https://www.the4thwheel.com/event-list
About gLOCAL Evaluation Week
An annual global knowledge-sharing event that brings together evaluators, researchers, governments, NGOs, funders, academic institutions, and development practitioners — one of the largest global platforms for dialogue on evidence, learning, accountability, and social impact.
About 4th Wheel Social Impact
Founded in 2010, 4th Wheel works with funders, NGOs, and government partners across India and South Asia on impact strategy, MEAL systems, and evaluation. We focus on building evidence systems that are field-grounded, methodologically rigorous, and useful to the people making decisions.
Warmly,
4th Wheel Social Impact Team
I cannot attend this presentation, but I am attending a few this week. Thank you for sharing!