Public attitudes to AI have shifted sharply in just a few years, from ambivalence to excitement to, increasingly, active dislike. That shift deserves to be taken seriously, not managed away.

In this episode of The PSC In Conversation, part of our The Next 20 Series, Antonio and Shuying make the case for a genuine national conversations on AI.

They discuss:

- Why trust in AI is falling in the UK and US, while countries like Brazil and India remain optimistic, and what that splintering of views tells us.

- The biggest unanswered question of all: who benefits, and by how much? No government anywhere has the answer yet.

- Why the public often knows things policymakers don't: from red lines and risk thresholds to how change lands in real lives.

- The ladder of citizen participation, from consultation to full co-production, and lessons from the Timms Review with DWP, the Cabinet Office, UKRI, DARE UK and BridgeAI.

- How to design engagement that's meaningful rather than extractive: real trade-offs, real feedback loops, real influence on outcomes.

The honest starting point is uncertainty. General purpose technologies change everything, and there's no implicit reason they change things for the better without deliberate direction. The public's worries, priorities and red lines aren't obstacles to navigate, they're a vital evidence base.