Into the Impossible With Brian Keating

This episode is sponsored by Abacus AI agent: agent.abacus.ai/keating

I fed it Carlo's recent papers and asked for the argument, the counterargument, and the evidence that would support or refute each side. Some of the questions and animations in this interview came out of that. It runs the top frontier models in one place, turns a dense PDF into a research report or a presentation with charts, builds interactive dashboards off a spreadsheet, and takes API access if you want it wired into your own tools.

Plans start at $10 a month. My link gets you the first month for $7: agent.abacus.ai/keating

The physicist who built loop quantum gravity says the second law does not say what you were taught it says, and that freedom does not require breaking a single law of nature. Order builds itself. Your head is older than your feet. And neither one is the strangest thing here.

Carlo Rovelli founded the loop quantum gravity program and wrote the relational interpretation of quantum mechanics.

We cover: why oil separating from water is not a violation of entropy, what Galileo got wrong about the one discovery he is most famous for, the paper Einstein published denying gravitational waves, why a chess program that deliberates tells you what your own freedom actually is, and what Rovelli would build with ten years and unlimited money.

“To imagine that to be free you have to violate the laws of nature is stupid.” — Carlo Rovelli

CHAPTERS

00:00 Your head is older than your feet

00:48 Things left alone get ordered. Who ordered them?

07:00 Low entropy might be about you, not the universe

08:38 You can be in two cities at once, just not to you

11:02 Why start with Galileo when Einstein knew more?

19:34 Galileo misunderstood his own discovery

22:32 Einstein published a paper denying gravitational waves

25:08 The definition of time that explains nothing

26:50 Two clocks, two metres apart, two answers

31:06 Nothing gets recorded without something being lost

33:12 The chess program deliberates. So do you.

37:06 Can a machine have Einstein’s happiest thought?

39:52 Thirty minutes versus a month

44:10 Dark matter made of black holes the size of a hair

51:10 Ten years, unlimited money, a room full of junctions

Rovelli told me the second law does not say what I was taught it says. Order builds itself.
Transcript, bonus content, and my Monday M.A.G.I.C. Message: https://briankeating.com/yt

Three moments worth the rewatch:
07:00 The entropy you were taught is a point of view
26:50 Why your head is older than your feet
39:52 The thing Claude did in 30 minutes

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:
Carlo Rovelli website: https://www.cpt.univ-mrs.fr/~rovelli/

The Relational Universe (book): https://www.amazon.com/dp/0735216118

Carlo Rovelli on Twitter/X: https://x.com/carlorovelli

My books:
Losing the Nobel Prize (memoir): http://amzn.to/2sa5UpA
Think Like a Nobel Prize Winner: https://a.co/d/03ezQFu
Focus Like a Nobel Prize Winner: https://a.co/d/hi50U9U
Galileo’s Dialogue (first-ever audiobook): https://a.co/d/iZPi9Un

Twitter/X: https://x.com/BrianKeating
Substack: https://briankeating.substack.com
Blog: https://briankeating.com/blog
Audio-only: https://briankeating.com/podcast

#CarloRovelli #freewill #entropy #physics #quantumgravity #briankeating #intotheimpossible

Learn more about your ad choices. Visit megaphone.fm/adchoices

They Built the Quietest Place on Earth to Find Dark Matter

One event in 220 days. That is what the search for dark matter looks like right now.

In this episode I talk with Rick Gaitskell, professor of physics at Brown University and a leader on the LZ dark matter experiment, about what it takes to build a detector that can see almost nothing at all: seven tonnes of liquid xenon, a mile of rock in an old gold mine, and shielding so careful that the radon seeping out of ordinary concrete counts as a serious enemy. We get into why the field spent decades hunting the WIMP, what happens to a 3.4 sigma result when you look at it honestly, and why a single event, however beautiful, cannot tell you what you want it to tell you.

We also talk about the human side of a project like this. The grad student who waited nine years for data. The physicist who walked away from investment banking. The uncomfortable fact that something like ninety-five percent of research does not work out, and what keeps people going anyway.

And the line that sits under all of it: we only get one damn universe, and nature does not care whether your theory is beautiful.

Chapters

00:00 One event in 220 days

06:00 Why they named it a WIMP

13:00 Ninety-five percent of research fails

20:00 He left investment banking for this

23:00 You are less than five percent of the universe

30:00 The fifty-year-old idea nobody can kill

45:00 The grad student who waited nine years

58:00 A gold mine and the quietest place on Earth

1:06:00 The radon in your basement is the enemy

1:21:00 The assumption that was hiding the signal

1:32:00 The event, and why they will not call it blinded

1:40:00 How 3.4 sigma becomes 2.6

1:49:00 The fog nobody can scrub

1:56:00 The dark matter we detected in 1956

My books

Losing the Nobel Prize (memoir)

Think Like a Nobel Prize Winner

Focus Like a Nobel Prize Winner

Galileo's Dialogue (first-ever audiobook)

Socials

Twitter/X: https://x.com/BrianKeating

Substack: https://briankeating.substack.com

Blog: https://briankeating.com/blog

Audio-only: https://briankeating.com/podcast

#intotheimpossible #briankeating #RickGaitskell #darkmatter #physics #AI #emergence #podcast
Learn more about your ad choices. Visit megaphone.fm/adchoices

For 50 years I taught my students that dark matter surrounds the Milky Way in a giant halo, or that Newton’s laws break down at low accelerations. The Gaia satellite just measured something that makes both of those answers uncomfortable. Subscribe if you want science with evidence, not speculation.

Gaia has measured over a billion stars. What its third data release shows is that the outer rotation curve of our own galaxy appears to decline in a Keplerian fashion. That should not happen if a dark matter halo is doing what we think it does. It also should not happen if MOND is right, because MOND produces flat curves for isolated galaxies. Both camps predicted something flatter. Neither got it.

I walk you through the physics from Kepler to Vera Rubin to Gaia DR3, show you why systematic errors from asymmetric drift, the Large Magellanic Cloud, and the Sagittarius dwarf galaxy matter enormously before anyone draws a conclusion, and lay out the three scenarios that still survive. The most interesting question here is not which side lost. It is why both sides were expecting the wrong curve from our own galaxy.

What you’ll hear:

-Why the neutrino technically counts as dark matter and why that still leaves most of the problem unsolved

-What Kepler’s first law of fatherhood has to do with galactic dynamics

-Why Stacy McGaugh and Mordehai Milgrom might be celebrating too soon

-What Cepheids, stellar streams, and globular clusters could tell us that Gaia cannot

-Whether the Milky Way is too messy to ever give us a clean answer

“The question is not which camp won, but why we were both expecting the wrong curve.” — Brian Keating

CHAPTERS

00:00 Two solutions. One problem. Neither wins.

02:12 What Vera Rubin actually found

03:18 We're inside the galaxy. That's the problem.

04:24 The asymmetric drift problem

05:04 What Gaia DR3 actually shows

06:34 Why the mass estimate is model-dependent

08:04 MOND: no radius, no problem. Until now.

08:56 Gaia may be saying both sides are wrong

09:24 Flaws lead to new laws

10:22 LMC, Sagittarius, and the galactic warp

12:28 Three viable explanations

12:42 Dark matter adjusts. MOND adjusts. But is that enough?

13:14 What could settle this

14:04 Are we expecting the wrong curve?

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

Related episodes:

Stacy McGaugh episode: https://www.youtube.com/watch?v=NnpyFk2WlME

My books:

Losing the Nobel Prize (memoir): http://amzn.to/2sa5UpA

Think Like a Nobel Prize Winner: https://a.co/d/03ezQFu

Focus Like a Nobel Prize Winner: https://a.co/d/hi50U9U

Galileo’s Dialogue (first-ever audiobook): https://a.co/d/iZPi9Un

Twitter/X: https://x.com/BrianKeating

Substack: https://briankeating.substack.com

Blog: https://briankeating.com/blog

Audio-only: https://briankeating.com/podcast

#darkmatter #MOND #MilkyWay #cosmology #briankeating #intotheimpossible #physics
Learn more about your ad choices. Visit megaphone.fm/adchoices

Did OpenAI just solve Navier–Stokes—and what would that actually mean for AGI?

Emad Mostaque joins Brian Keating LIVE to examine the claimed breakthrough, what would count as a verified solution, and whether AI solving a major mathematical problem changes the debate about intelligence—or control.

Watch my explainer about AI and the Navier-Stokes problem, including insights from Terry Tao https://youtu.be/je9X59psm6U?si=wGbcLhXmf47jeCHL?sub_confirmation=1

A convincing answer isn’t necessarily a proof. A mathematical breakthrough isn’t automatically AGI. But if this claim holds up, how much should we update our expectations?

Emad helped bring Stable Diffusion to the world and signed the letter calling for a pause on frontier AI development. In his previous conversation with Brian and AI safety researcher Roman Yampolskiy, he questioned how such a pause could work—and highlighted swarm intelligence as a particularly dangerous and unpredictable risk.

Now we return to the question underneath the excitement: if AI’s capabilities are accelerating, is our ability to understand and control them keeping up?

ON THE TABLE:

• What exactly is being claimed about Navier–Stokes?
• What separates a proposed solution from an accepted mathematical proof?
• If the result holds up, does it demonstrate AGI—or something narrower?
• Can AI produce discoveries that humans struggle to verify?
• What would this change about open-source AI and the argument for a pause?
• Are we watching the wrong risk: a single superintelligence, or swarms of AI agents?
• What evidence would change Emad’s mind?

Bring your questions. We’ll examine the claim, the evidence, and what follows if it’s right.

This livestream discusses a claimed result; it does not assume the Millennium Prize Problem has been resolved.

FEATURED GUEST:

Emad Mostaque on Twitter/X: https://x.com/EMostaque
I.I.I. Inc.: https://ii.inc/
Stable Diffusion / Stability AI: https://stability.ai/

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? Claim your meteorite: https://BrianKeating.com/edu

Subscribe: https://www.youtube.com/DrBrianKeating?sub_confirmation=1

Support Into the Impossible on Patreon for unfiltered bonus content and live monthly Office Hours: https://www.patreon.com/drbriankeating

Join this channel for perks, monthly Office Hours, and your name in the Member Roster: https://www.youtube.com/channel/UCmXH_moPhfkqCk6S3b9RWuw/join

MY BOOKS:

Losing the Nobel Prize: http://amzn.to/2sa5UpA
Think Like a Nobel Prize Winner: https://a.co/d/03ezQFu
Focus Like a Nobel Prize Winner: https://a.co/d/hi50U9U
Galileo’s Dialogue audiobook: https://a.co/d/iZPi9Un

CONNECT WITH BRIAN:

Twitter/X: https://x.com/BrianKeating
Substack: https://briankeating.substack.com
Blog: https://briankeating.com/blog
Audio-only: https://briankeating.com/podcast

#OpenAI #NavierStokes #AGI #EmadMostaque #ArtificialIntelligence #BrianKeating #IntoTheImpossible
Learn more about your ad choices. Visit megaphone.fm/adchoices

The man who open-sourced the most-used AI image model in history sat down with the man who has spent a decade proving superintelligence cannot be controlled. Brian set up a debate. What emerged was something more unsettling than any debate.

Roman Yampolskiy is the computer scientist who coined the term AI safety and author of AI: Unexplainable, Unpredictable, Uncontrollable. Emad Mostaque is the co-founder of Stable Diffusion and the only AI CEO who signed the pause letter. He now says he doesn’t know how a pause could work. Yampolskiy thinks that is the only option left.

The question underneath everything is simple: if you have a 50% chance of wiping out civilization and you build it anyway, what are you actually doing? We cover what AI safety researchers actually think the danger is, why nobody has published a paper, filed a patent, or shipped a prototype for controlling a superintelligence, what the Qwen weights being out means for the pause argument, and why Mostaque thinks swarm intelligence is the most dangerous and most unpredictable risk vector we have.

What you’ll hear:
-Why both guests think P(doom) tells you less than you’d hope
-What it means that no company, no lab, and no team has a patent on controlling superintelligence
-Why the models the public receives are slightly lobotomized
-The difference between an AI swarm and the ASI everyone is debating
-Why giving every psychopath access to a cutting-edge intelligence weapon is incoherent safety strategy
-What it would actually take to update Yampolskiy’s assessment

“We either do it, or we die. There is nothing for you to gain by doing it.” — Roman Yampolskiy

CHAPTERS
00:00 A debate that wasn't a debate
00:36 Turing test, AGI, superintelligence: where are we?
02:02 The open source argument nobody wins
03:32 Pause frontier AI forever. Which button?
05:04 P(doom): parameterizing our ignorance
09:02 Nukes are inefficient. AI isn't.
11:46 The lobotomized model problem
18:04 Decade-old problems solved weekly now
26:18 We either do it or we die
31:10 Stop the training or stop the funders
33:34 Lipstick on a Shoggoth
35:50 No paper. No patent. No framework.
39:42 Train only on what you need
44:14 What lowers Mostaque's p(doom)?
52:00 Same future. Two perspectives.

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 Guests:
Roman Yampolskiy on Twitter/X: https://x.com/romanyam?lang=en
AI: Unexplainable, Unpredictable, Uncontrollable (book): https://www.amazon.com/dp/103257626X
MIRI: https://intelligence.org

Emad Mostaque on Twitter/X: https://x.com/EMostaque
I.I.I. Inc.: https://ii.inc/
Stable Diffusion: https://stability.ai/

My books:
Losing the Nobel Prize (memoir): http://amzn.to/2sa5UpA
Think Like a Nobel Prize Winner: https://a.co/d/03ezQFu
Focus Like a Nobel Prize Winner: https://a.co/d/hi50U9U
Galileo’s Dialogue (first-ever audiobook): https://a.co/d/iZPi9Un

Twitter/X: https://x.com/BrianKeating
Substack: https://briankeating.substack.com
Blog: https://briankeating.com/blog
Audio-only: https://briankeating.com/podcast

#AIrisk #aisafety #stablediffusion #RomanYampolskiy #EmadMostaque #briankeating #intotheimpossible
Learn more about your ad choices. Visit megaphone.fm/adchoices