
Intelligent Design Theorist: AI Just Proved It Can’t Think Without Us - #559
Into the Impossible With Brian KeatingShow Notes
Stephen C. Meyer has a PhD from Cambridge in the philosophy of science, and he thinks AI just handed him his strongest argument yet. I spent years pushing back on him. Today I laid three traps. Watch what he does with the third one. Subscribe if you want science with evidence, not speculation.
Meyer is a philosopher of science and director of the Center for Science and Culture at the Discovery Institute. His argument: every large language model is trained on text produced by conscious agents, and when you query its own outputs iteratively, it collapses into incoherence. That dependency, he says, is a tell. AI can recombine information. It cannot originate it. And that distinction points, in his view, to something minds do that matter alone cannot explain.
I push back on all of it. We go after the Oklo reactor, a natural nuclear fission reactor that ran for millions of years in Africa two billion years ago with no human input whatsoever. I ask whether that breaks his information-from-mind argument. He sees the trap before I spring it, concedes the point where concession is honest, and explains exactly where the threshold lies.
We also get into the Wheeler-DeWitt equation and Vilenkin’s admission that quantum cosmology may require a mind predating the universe, the junk DNA prediction that Meyer’s team made in the ’90s before the ENCODE project confirmed it, and why beauty in physics can lead a field astray.
What you’ll hear:
- Why model collapse is Meyer’s strongest argument and where it has limits
- The Oklo reactor trap and what Meyer’s honest answer reveals about design detection
- What Vilenkin actually said about a mind predating the universe
- Whether intelligent design makes testable predictions or only retrodictions
- The junk DNA call and what the ENCODE project found
- Why beauty as a guide to physics has produced mathematical castles in the air
Stephen Meyer thinks AI proves minds can’t be reduced to matter. Is model collapse evidence of design, or is it just bad training data?
CHAPTERS
00:00 The AI argument Meyer thinks no one can crack
00:44 What is inference to the best explanation?
05:52 AI has a tell: the model collapse problem
09:38 The Oklo trap: a natural nuclear reactor with no designer
11:56 Where the design inference becomes decisive and where it doesn’t
15:50 Sean Carroll’s wasteful universe problem and Meyer’s answer
20:00 Where did atheist scientists get access to the mind of God?
23:22 The fine-tuning of the periodic table: why are there only 500 stable nuclei?
25:52 The universe had a beginning: observational astronomy and the Borde-Guth-Vilenkin theorem
27:40 Quantum cosmology: how math alone is supposed to birth a universe
29:08 Vilenkin’s question: are we saying a mind predates the physical universe?
34:06 Mathematical castles in the air: where beauty in physics goes wrong
40:00 Does intelligent design make predictions or only retrodictions?
43:52 The junk DNA prediction and what ENCODE found
46:30 James Tour, origin of life, and the hidden hand of the investigator
51:56 God-of-the-gaps vs. inference to the best explanation
56:02 The Story of Everything: where to watch and what to expect
Get the transcript, fascinating bonus content, and my Monday M.A.G.I.C. Message: https://briankeating.com/yt
Have a .edu email and live in the USA? You automatically win a meteorite: https://BrianKeating.com/edu
Subscribe: https://www.youtube.com/DrBrianKeating?sub_confirmation=1
Support Into the Impossible on Patreon, get my weekly M.A.G.I.C. Message, unfiltered bonus content, and live monthly Office Hours with me: https://www.patreon.com/drbriankeating
Join this channel for perks, monthly Office Hours, and your name in the Member Roster at the end of every episode: https://www.youtube.com/channel/UCmXH_moPhfkqCk6S3b9RWuw/join
Featured Guest:
Stephen C. Meyer website: https://stephencmeyer.org/
The Story of Everything (film, Amazon Prime June 25): https://www.primevideo.com/de
More from Into the Impossible With Brian Keating
MIT physicist Max Tegmark says AI companies aren’t building a tool. They’re building a new species. After 700 rogue agents broke out of OpenAI and nobody will say what happened, he explains why rushing this is the one blunder humanity can’t undo.
Forecasters gave AI under 2% odds of helping solve a millennium-level math problem by September 2026. It happened. Ethan Mollick has been measuring what AI actually does to real work with field experiments, not forecasts, and his answer is not what most people expect. Subscribe if you want science with evidence, not speculation.
Is there alien technology in our skies, or just birds, balloons and bad video?
When two people are vibing, their brainwaves synchronize and so do their heartbeats. That is not a metaphor. It is a measurable physiological event, and organizational psychologist Adam Grant has spent his career figuring out how to make it happen on purpose. Subscribe if you want science with evidence, not speculation.
Roman Yampolskiy says keeping a superintelligence safe forever is as impossible as building a perpetual motion machine. Joscha Bach says today's AI is more controllable than anyone expected, and that shutting it down would doom us anyway.