Connect your Claude, Codex or OpenClaw for full transcripts on millions of podcasts. Two week free trial, cancel anytime.

Start free trial →
Ja Jim al-Khalili On Into the Impossible With Brian Keating

“However smart LLMs are, and despite the likes of my good friend Richard Dawkins saying that he believes that Claude is truly conscious, I think he's very, very wrong. I don't think that means that AI wouldn't one day become sentient and self-aware.”

Into the Impossible With Brian Keating · Ideas & Essays · October 2026

“However smart LLMs are, and despite the likes of my good friend Richard Dawkins saying that he believes that Claude is truly conscious, I think he's very, very wrong. I don't think that means that AI wouldn't one day become sentient and self-aware.” — Jim al-Khalili, Into the Impossible With Brian Keating

Context

Brian Keating asked whether AI can experience time. Al-Khalili argued that experiencing time requires a sense of self, which AI is a long way from; it may one day become self-aware without feeling joy or envy the way humans do.

Transcript

Into the Impossible With Brian Keating Around 38:44 into the episode
Brian Keating

…that. But my question is, can they experience time? Do they experience time? They don't have psychologies. They have some ways of representing memory, but often it's quite bad. Can artificial consciousness participate in the entanglement that you maybe potentially are beginning to subscribe to?

Speaker 1

Instagram teen accounts come with automatic protections that limit who can contact teens, the content they can see and the time they spend on the app. Learn more about teen accounts and Instagram's ongoing work to protect teens online at instagram.com slash teen accounts.

Jim al-Khalili

…is accessing memories in our brains with where we've stored a record of events that have happened in the past. But I think the notion of being of experiencing time moving, anticipation of the future, remembering the past, being aware that we live in the present moment, I think that requires a sense of knowing oneself, the introspection, sentience. Artificial intelligence is a long way from that. However smart LLMs are, and despite the likes of my good friend Richard Dawkins saying that he believes that Claude is truly conscious, I think he's very, very wrong. I don't think that means that AI wouldn't one day become sentient and self-aware. I think that may not be thinking. Like a human being will think may not feel joy and anger and envy or whatever in the same we do, but it might become self-aware.

Brian Keating

Very good, Jim. Oh, last question. You became a grandfather, I think, since we last spoke. Is that correct? I think. Has time taken on a different meaning since that happened? Is it qualitatively different since becoming a father? You know, they say that grandfathers and grandchildren get along…

Jim al-Khalili

I think in lots of ways. And also, just getting older. I remember when I turned 40, that was great. Well, no problem with that. You know, I turned 50. I had a big fabulous party. I turned 60 and you think, my God, this is getting serious now. So there's that notion of I've lived more life than I…

Brian Keating

I can't wait for you to do many more diaper changes, nappy changes, but also many more books. And hopefully we'll get to meet up someday.

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.

More from Into the Impossible With Brian Keating