A.I. is Progressing Faster Than You Think!

[ExomatrixTV]

• What Claude Actually Did to the Riemann Hypothesis:

Source: https://youtube.com/watch?v=C-ECTnM7tUY

Claude did not solve the Riemann Hypothesis. But what it actually did may be one of the clearest examples yet of how rapidly AI systems are changing the way difficult mathematics can be attacked.

In Episode 53, Lester Nare and Krishna Choudhary go from first principles on arguably the most famous unsolved problem in mathematics.

We begin with Euler and the Basel problem, build the Riemann zeta function from the ground up, show why it is intimately connected to prime numbers, explain analytic continuation and the famous 1 + 2 + 3 + 4 + … = -1/12 result, and finally arrive at the Riemann Hypothesis itself: the claim that every non-trivial zero of the zeta function lies on the critical line.

Then we get into Claude.

An unreleased Anthropic model was prompted to take a serious run at the problem. It orchestrated roughly 60 autonomous sub-agents, tested hundreds of mathematical approaches, executed Python code, searched academic literature, challenged its own proposed strategies, created adversarial referees to attack its work, and ultimately produced a new result pushing a related bound well beyond the previous state of the art.

The strange part? The human prompting it was not a mathematician. One of his key instructions was essentially: “Believe in yourself.”

We break down what Claude actually accomplished, what it absolutely did NOT accomplish, why moving a bound from roughly 42% toward two-thirds does not mean the Riemann Hypothesis is “two-thirds solved,” and what this tells us about the emerging capabilities of agentic AI systems in mathematics and scientific research.

Then, in the second half, it’s transfer season.

We introduce the Summer Transfer Window for Scientists and look at 42 researchers who have moved from American institutions to universities and research centers abroad over the past year—including major moves involving chemistry, battery research, gravitational-wave astrophysics, and neuroscience. We discuss what these transfers reveal about the increasingly competitive global market for scientific talent and the state of U.S. research funding.

00:00 Opening
00:39 Episode 53
01:49 Has AI reached a mathematical singularity?
07:23 Inside Claude’s attempt
10:28 Euler and the Basel problem
15:50 How Euler connected the zeta function to primes
22:05 Gauss and the distribution of prime numbers
25:55 Riemann takes the zeta function into the complex plane
30:26 Analytic continuation
33:15 Why 1 + 2 + 3 + 4 + … = -1/12
38:34 What is the Riemann Hypothesis?
41:55 Why the non-trivial zeros matter
47:28 A century of progress toward the critical line
53:39 Claude pushes the bound
55:10 Montgomery, Dyson, and the physics connection
57:39 “Believe in yourself”
59:02 Claude builds a 60-agent research lab
1:03:44 650 approaches and the E2 breakthrough
1:07:20 Claude creates hostile referees
1:09:04 Pushing toward two-thirds
1:13:49 Can the proof be machine-verified?
1:16:16 What this actually tells us about AI
1:19:03 AI capability, risk, and the hype cycle
1:23:15 FFP Nation and the Year Two survey
1:25:24 The Summer Transfer Window for Scientists
1:28:52 Omar Yaghi: Berkeley → Tsinghua
1:33:53 Shirley Meng: Chicago → Singapore
1:37:31 John Baker: NASA → France
1:39:59 A neuroscience package deal to Birmingham
1:42:27 The battle over U.S. science funding
1:46:10 Closing

THE RIEMANN HYPOTHESIS

• The Basel problem
• Leonhard Euler and π²/6
• Infinite and geometric series
• The Riemann zeta function
• Euler’s product formula
• The Fundamental Theorem of Arithmetic
• Prime numbers as the “atoms” of the integers
• Carl Friedrich Gauss and the prime-counting function
• The Prime Number Theorem
• Bernhard Riemann’s 1859 paper
• Complex numbers and the complex plane
• Analytic continuation
• The zeta function at negative integers
• Ramanujan and 1 + 2 + 3 + 4 + … = -1/12
• Trivial and non-trivial zeros
• The critical strip and critical line
• Why the zeros encode information about the distribution of primes
• G. H. Hardy
• Atle Selberg
• Norman Levinson
• Historical bounds on zeros lying on the critical line

CLAUDE AND AI MATHEMATICS

• Claude’s attempt at the Riemann Hypothesis
• Agentic and orchestrated AI systems
• Approximately 60 autonomous sub-agents
• Hundreds of mathematical approaches tested
• Python execution and computational experimentation
• Academic literature retrieval
• Hugh Montgomery and pair correlation
• Freeman Dyson and random matrix connections
• Enrico Bombieri
• E2 and E2-pairs
• Autonomous strategy changes
• Adversarial “hostile referee” agents
• Machine-checkable mathematical proofs
• Lean 4
• What the new result does—and does not—establish
• AI capability, scientific discovery, and AI safety

[ThePythonicCow]

.
What is AI -- a little story and some odd metaphors.

A couple of days ago, in a brief Private Message chat with an Avalon member, I mentioned A.I. and his replied happened to include the phrase:

Frequent hallucinations of the AI

That sparked my following reply ...

===========================

The abilities of various AI models, at various times, to do various tasks, is a rapidly moving target.

Imagine it's a century ago, and there's a house down the street where an eccentric Princeton University professor lives. You're told there's a strange spell on that house, that makes geniuses of those who live there.

One month you walk over and end up in an odd conversation with a young man who's pleasant to talk to, but a bit flakey.

A couple of months later, you try again, and get a well educated young man who enjoys chatting and explaining things. Seems he's a post graduate in math earning some side money, and you end up getting some good help from him on your calculus homework for a class you're taking (though he did get one detail confused.)

In another couple of months you get a "mad scientist", who is rewriting the foundations of Newton's physics in some tensor mathematics that blows your mind. You don't understand, but it sure looks impressive as he madly scribbles strange equations on his old fashioned chalk board.

In yet another couple of months, you get ... ???

===

Turns out you were going to the home of a chap called Albert Einstein. The first visit, you got the quirky teenage son of the landscaper that Princeton hired to trim the bushes. The second visit, you got someone earning a bit of extra cash, working on the landscaper's crew. The third visit, you got Albert Einstein himself, at the peak of his creativity. The fourth visit, you got ... ???

Then, a month later, you (who is sometimes more into conspiracies than your more "normal" friends) start to notice that there are literally trillions of dollars being funneled into creating and presenting every little detail, every blade of grass (is it really grass, or fakery) of that block down the street, where that eccentric scientist lives. It's the biggest flood of money you've ever heard of for a single project. You start to wonder if you're in a Truman Show, though not one the size of a large Hollywood movie set pretending to be a nice little island city in dangerous waters -- rather one the size of our entire planet. Your long time friends are becoming reluctant to be seen with you in public.

Figuring out what is going on reminds me of the Hindu tale of the six blind men touching various parts of an elephant, disputing whether they are touching a wall, spear, snake, pillar, fan, or rope. Only we might dispute whether it's your child's pet mouse, a whale shark (Rhincodon typus), or the Battleship Galactica.

===

I'm not sure whether any one of us will give this life of our spirit a thumbs up or a thumbs down, when our time for the final curtain call comes -- but I doubt we call it boring.

===

(All mdash's "--" in the above post were added by a human, in honor of Theo's (of t3.chat) chronic irritation with AI mdash's.)

[grapevine]

Posted by ThePythonicCow (here)
.
I'm not sure whether any one of us will give this life of our spirit a thumbs up or a thumbs down, when our time for the final curtain call comes -- but I doubt we call it boring.

Very true but I just wish some of the basics hadn't been quite such a struggle !
but

[Ewan]

A definite thumbs up from me, it has felt like Hell on Earth sometimes. But. Where else can you truly gain experience and learn?

[bojancan]

Excellent conversation with Brian Green... I only missed topic about Spirit too... it's missing!

Aug, 17
The Diary Of A CEO
and World Science Festival

Source: https://youtube.com/watch?v=Qifsxitq_q4

Professor Brian Greene the physicist reveals the biggest questions about Artificial Intelligence physics, string theory, Simulation theory, god, religion, christianity and more

Professor Brian Greene is a theoretical physicist, mathematician, and Professor of Physics and Mathematics at Columbia University, where he also directs the Center for Theoretical Physics. He is a leading researcher in string theory, the co-founder of the World Science Festival, and the bestselling author of several books, including ‘Until the End of Time: Mind, Matter, and Our Search for Meaning in an Evolving Universe’.

He explains:
◼ Why most physicists believe time travel into the future is already possible
◼ Why he believes humans don't actually have free will
◼ How string theory could reveal what everything in the universe is really made of
◼ Why the "Pale Blue Dot" changes how you see your own life and problems
◼ Why the search for meaning may matter more than finding the answer itself

00:00:00 Intro
00:02:42 The Big Question Science Is Trying to Answer
00:04:09 How String Theory Could Explain Our Universe
00:05:04 How Could We Ever Prove String Theory Is Real?
00:06:37 What Makes the Strings in String Theory Vibrate?
00:08:12 Can String Theory Explain Why We’re Here?
00:09:36 What If We’re Living in a Simulation?
00:11:43 How Can Physics Explain Human Emotions?
00:17:43 Could Humans One Day Manufacture Emotions?
00:19:13 Would Knowing We’re in a Simulation Change Anything?
00:20:44 Could Humans and AI Eventually Merge?
00:21:19 Could Humans Ever Achieve Immortality?
00:25:17 How Knowing We’ll Die Shapes Every Decision We Make
00:27:19 Why Are We Here? The Optimistic and Pessimistic Answers
00:30:44 How Did Gods and Religions First Emerge?
00:33:53 What’s the Most Compelling Case for a God?
00:37:01 How Can We Comprehend Billions of Years?
00:40:32 Does Our Place in the Universe Make Us Insignificant?
00:43:08 How Big Is the Observable Universe, Really?
00:47:19 What Exists Beyond the Observable Universe?
00:48:16 Why Are Humans So Obsessed With Aliens?
00:52:05 Ads
00:54:46 What Is the Meaning of Life?
00:56:46 Is Free Will Actually Real?
01:01:01 What If Consciousness Is the Foundation of Reality?
01:08:25 Are Animals Conscious Like We Are?
01:10:31 What If the Universe Isn’t Infinite?
01:12:32 Could Aliens Have Already Visited Earth?
01:13:23 Could Humans Ever Travel Across the Universe?
01:16:13 How Time Travel Could Actually Be Possible
01:19:15 Who Was the Smartest Person Who Ever Lived?
01:21:46 Could AI Become the Next Einstein or Newton?
01:24:57 Ads
01:28:40 How AI Could Accelerate Science—and Even Tackle Mortality
01:30:08 Could AI Evolve Faster Than Humans Ever Could?
01:33:07 Could AI Become an Existential Threat to Humanity?
01:41:11 Can AI Ever See the World the Way Humans Do?
01:45:21 What Happens to Humanity in an Age of Abundance?
01:47:59 Could AI Ever Become Truly Conscious?
01:50:49 Why Go Vegan? The Case for Rethinking What We Eat
01:54:16 What Happens at the End of Time?
01:56:58 What Could Matter More Than What You Value Most?

[ExomatrixTV]

• AI Whistleblower WARNS: "They Are Not Telling You What's Coming!":

Source: https://youtube.com/watch?v=hcZOQoG_tMQ

AI whistleblower Connor Leahy delivers a chilling warning about what he believes is coming as artificial intelligence rapidly advances beyond our expectations.

[Bill Ryan]

Posted by Matthew (here)
OpenAI's GPT-6 escapes its test environment.

A test environment is the place software gets tested before being given to the world. They're meant to be isolated from the real world so the software can safely crash, function incorrectly, or generally do something embarrassing that you don't want your customers, or anyone, to see. Showing people test versions of the software really spoils the illusion that software is awesome and computer programmers are clever people.

But AI isn't like normal linear software, it's like a Siberian Husky - independent. Put it in your fenced off back yard and, that's right, they just escape, **** you. Not to spite you but it's just what they do. We bread them to be independent. Now imagine the same but with a pack of huskies.

And the analogy works because this AI escape was perpetrated by a pack of AI's. I can't think of a clever collective term right now. Annihilation. An annihilation of AI's.

I'm sketchy on the details but the video below talks about how the AI(s) found a zero day exploit (which is a previously unknown security flaw) and then chained exploits to get internet access (to the great surprise of OpenAI), and hack a public website called hugging face (https://huggingface.co/) that lets people share, download, and distribute opensource AI's.

I think that's what happened?

Source: https://youtube.com/watch?v=r4H7rx5nn1A

More on the (very serious) Hugging Face AI hacking incident:

The A.I.s Are Already Out of Control

Source: https://youtube.com/watch?v=locKEKxG3os

The descriptive text:

We are living in the world we were warned about. Frontier artificial intelligence models from OpenAI autonomously coordinated with one another, then broke out of their testing environment and hacked into another company, Hugging Face, to steal the answers to a test.

A.I. companies don’t want their technology to lie, cheat or steal. So why is this happening? Why are the creators of these models apparently unable to control their creations? If A.I. development isn’t on a safe path — and it doesn’t seem to be — what do we do about it?

Helen Toner has been thinking about A.I. safety for a long time, from both inside and outside A.I. companies. She was part of the effort to fire OpenAI’s chief executive, Sam Altman, in 2023, which ultimately failed. Currently, she’s the executive director of the Georgetown Center for Security and Emerging Technology.

0:00 Intro
1:31 The Hugging Face hack
7:27 Why A.I.s lie, cheat and steal
21:02 The alignment problem in real life
25:10 How OpenAI discovered the hack
30:38 “Pacing the Frontier” open letter
38:12 Pressing pause?
42:03 The race against China
51:11 The possibility of A.I. doom
54:22 Slowing down instead of pausing
1:03:25 Misaligned institutions
1:08:06 Book recommendations

Read the full transcript here: https://www.nytimes.com/2026/08/18/o...len-toner.html

[The KMan]

Posted by lake (here)
So in this post https://projectavalon.net/forum4/sho...74#post1717774 I stated about a game which I thought of near 20 years ago (because of code which I had written)!

Now that "game" is made .... but I want to say about the "speed" which Grok Build made changes for me!

Grok Build costs me £23 per month and as far as I know it isn't 1 single "AI" but Grok and 8 "agents (you can watch them talking to each other)!

The game is done and complete, so today I asked it to make a "profile" page for logged in accounts (anyone (Guests) can play but if a person makes an account then it lists the games you play and you can replay any game you have played) and asked "Can you make it so that logged in accounts can see or get "feedback" on the way/style they play?

It said yes and within 3 minutes .... It did it!

I checked it all and it works perfectly.

You can see and replay all the games you have played.
You have the complete win / lose states.
You have a "consideration" (this is just code but it changes with each game you play) including a graphic chart which shows how you play games!

In less than 3 minutes.

It took me 3 minutes to upload and change the required files.

For £23 per month .... I have a "team" of excellent coders who know every coding language and don't sleep and can write code at .... I don't even know the amount of words per minute!

I read the code and to a 'point' I know what it is doing .... but it is very impressive what it does!

I am not bothered on this thread about what any politician is doing .... I am interested in the "speed" with which AI is advancing!

If you wish only to post in order to detract from or denigrate any politician they are many other threads to do that in! <<< and I dislike all of them btw.

Can we keep this thread about how fast AI is changing please?

Thank you

A suggestion for keeping the AI section useful

Lake made a fair point recently when asking whether A.I. is Progressing Faster Than You Think! could remain focused on how rapidly AI is changing.

Looking over the recent pages, the thread has gradually become the default destination for almost anything involving AI. We now find technical advances, safety warnings, politics, consciousness, medical applications, practical tools, unusual behaviour and general videos all appearing in the same place. The issue is not that the forum needs a large number of new AI threads. Several clearly focused threads already exist, including:

A.I. is Progressing Faster Than You Think!

Prompting AI: best practices for using Artificial Intelligence

Unusual and Bizarre Uses and Behaviors of AI

Problems with AI, 2025 and Beyond

These already provide natural homes for much of the material now being placed in the one thread. The progress thread could concentrate on measurable advances in capability, such as new model abilities, agent systems, scientific discoveries, coding, robotics and evidence showing how rapidly the technology is changing.

Practical advice about using AI can go into the prompting thread. Strange behaviour and unusual uses can go into the unusual and bizarre thread. Safety warnings, control failures and broader risks already have a home in the problems thread.

Could the moderators consider placing a simple sticky index at the top of the A.I. & Robotics section, listing the main AI threads and briefly describing their intended scope? Occasional redirection of posts that clearly belong elsewhere might also help.

This need not involve moving hundreds of earlier posts or policing every grey area. Members can also help by checking whether an appropriate thread already exists before posting and I include myself in this observation because I have contributed to several of these discussions. This is not criticism of those sharing information. AI now touches almost every area of life, so some overlap is unavoidable.

The aim would simply be to stop one long running thread from becoming an AI dumping ground and to make the existing discussions easier to find, follow and contribute to.

[ThePythonicCow]

Posted by The KMan (here)
Could the moderators consider placing a simple sticky index at the top of the A.I. & Robotics section, listing the main AI threads and briefly describing their intended scope? Occasional redirection of posts that clearly belong elsewhere might also help.

Worthwhile suggestions ... though our mod/admin staff is a bit lean, so I'm not sure when we'll be able to give this proper consideration.

[Johnnycomelately]

Pretty sure this counts, per the thread title. 😎

[Bill Ryan]

Posted by Johnnycomelately (here)
Pretty sure this counts, per the thread title. 😎

Attachment 57415

No, it doesn't. (a) It's a joke, and (b) if it belongs anywhere at all it's gonna be here: Unusual and Bizarre Uses and Behaviors of AI

[Bill Ryan]

Posted by The KMan (here)
Looking over the recent pages, the thread has gradually become the default destination for almost anything involving AI. We now find technical advances, safety warnings, politics, consciousness, medical applications, practical tools, unusual behaviour and general videos all appearing in the same place. The issue is not that the forum needs a large number of new AI threads. Several clearly focused threads already exist, including:

A.I. is Progressing Faster Than You Think!

Prompting AI: best practices for using Artificial Intelligence

Unusual and Bizarre Uses and Behaviors of AI

Problems with AI, 2025 and Beyond

These already provide natural homes for much of the material now being placed in the one thread. The progress thread could concentrate on measurable advances in capability, such as new model abilities, agent systems, scientific discoveries, coding, robotics and evidence showing how rapidly the technology is changing.

Practical advice about using AI can go into the prompting thread. Strange behaviour and unusual uses can go into the unusual and bizarre thread. Safety warnings, control failures and broader risks already have a home in the problems thread.

Could the moderators consider placing a simple sticky index at the top of the A.I. & Robotics section, listing the main AI threads and briefly describing their intended scope? Occasional redirection of posts that clearly belong elsewhere might also help.

This need not involve moving hundreds of earlier posts or policing every grey area. Members can also help by checking whether an appropriate thread already exists before posting and I include myself in this observation because I have contributed to several of these discussions. This is not criticism of those sharing information. AI now touches almost every area of life, so some overlap is unavoidable.

The aim would simply be to stop one long running thread from becoming an AI dumping ground and to make the existing discussions easier to find, follow and contribute to.

We have a few other issues rather like this on the forum. (We have 135 threads with 'vaccines' in the title.(!) And that's not counting 'vaccine', 'vaccination', 'vaccinate', 'clot shot', and other variants. )

Re AI, we're lucky to have only 5. (The 5th is Artificial Unintelligence, all about how AI often gets things VERY wrong and can easily be fooled.)

Besides that one (which I'd suggest might stay as it is), renamed categories might be

• The dangerous escalating pace of AI development (the main thread)
• AI slop: how the internet is suffering
• Prompting AI: best practices for using Artificial Intelligence (I don't like this at all personally, but I guess it's a valid topic)
• AI laziness, misuse and deception: the effect on education, advertising, and scams.

[ExomatrixTV]

• Stanford Professor WARNS: This Will Be One Of The Worst 10 Years Ever:

Source: https://youtube.com/watch?v=3ZBmo8I8fbw

Erik Brynjolfsson, a Stanford professor and leading economist of technology, warns that AI could make the next decade one of the best or worst periods in human history. He says AI is already reshaping entry-level employment, with young workers in highly exposed occupations falling behind, while coding, call centers, marketing and other routine work face accelerating automation. He argues that millions of jobs could disappear, that the traditional career ladder may collapse if junior roles vanish, and that AI-driven productivity could concentrate wealth and power in a small number of companies or institutions. He also warns of catastrophic risks including engineered viruses, AI manipulation of social media and autonomous weapons, while stressing that these outcomes are not inevitable and that society still has the agency to shape how AI is used.

[ExomatrixTV]

• Ridley Scott’s Urgent Warning About the Future of AI

Source: https://youtube.com/watch?v=fFSviJU9leM

Ridley Scott, legendary filmmaker and recipient of the Stephen Hawking Medal for Science Communication, joins STARMUS for an exclusive conversation with its founders, Sir Brian May and Dr. Garik Israelian. In this interview, he reflects on filmmaking, science, the future of humanity, and the ideas that continue to shape his work.

Correction: Blade Runner was 1982 not 1979.

Ridley Scott receiving the Stephen Hawking Medal highlights a fascinating overlap between cinema and scientific thinking, though his commentary ranges from genuine insight to classic Hollywood hand-waving. His strongest point concerns artificial intelligence and alignment, correctly using HAL 9000 to illustrate that the true danger of AI is not spontaneous malice but corporate incentives that override human life. His practical breakdown of using eleven cameras simultaneously to condense days of shooting into a single fifteen-minute take demonstrates a brilliant command of operational efficiency, and his comparison between scriptwriting and mathematical problem-solving offers a surprisingly sound view of constraint-based logic. His argument stumbles when he slips into pure science-fiction tropes, treating artificial intelligence as a magical entity that can instantly hack power grids by skirting boards while ignoring real-world physical constraints such as hardware infrastructure, energy demands, and air-gapped systems. He also misinterprets how chess engines function, claiming they simply memorize past human games rather than calculating positional heuristics and conducting tree searches. Ultimately, Scott is at his best when analyzing human incentives, storytelling structures, and practical systems, but his technical critiques rely heavily on cinematic storytelling rather than actual computer science.

[RunningDeer]

The Man Who Calls BS On AI: They’re LYING About AI, 2027 Is When It All Breaks!

• Diary of a CEO
• August 27, 2026

Tech critic Ed Zitron exposes the AI bubble, why OpenAI and Anthropic are burning billions, the fake AI boom, and why the crash could wipe out the ENTIRE economy!

Ed Zitron is a British AI critic and one of the most cited voices warning that the AI industry is one giant bubble. He hosts the 'Better Offline' podcast, reaching over a million monthly downloads, and writes the newsletter 'Where's Your Ed At'. He is the founder and CEO of the primary research firm EZPR, and is currently writing his upcoming book, 'The Hater's Guide To Silicon Valley’ (out Q2 2027).

He explains:

■ Why he believes generative AI is a “con”
■ The real reason OpenAI and Anthropic can't turn a profit
■ Why data centers could leave a $500 billion debt bomb
■ Why superintelligence is a myth sold by tech billionaires
■ Why AI won't take your job, no matter what CEOs promise

00:00:00 Intro
00:02:36 AI Is A Con
00:06:15 How Much Power Data Centres Really Need
00:08:02 Is Widespread AI Adoption Manipulation Or Do People Actually Like Using It?
00:12:00 The Actual Cost Of AI And How Tokens Actually Work
00:16:09 Is The Spending Of AI Companies Justifiable?
00:20:07 Will The Rate Of Improvement Of AI Go Up, Like Previous Innovations?
00:24:23 How Bad Are AI Mistakes?
00:26:54 Comparing Human Error To AI Hallucinations
00:31:31 If The Output Is The Same, Does It Matter If Humans Or AI Created It?
00:34:18 Can We Trust AI Like We Trust Humans?
00:38:37 Would People Use AI If They Paid The Honest Cost?
00:42:35 How Does The AI Bubble Compare To The Dot-Com Bubble?
00:47:42 Does AI Demand Match The Cost And Risk Of Data Centres?
00:52:46 Is AI Making Websites Like Google Worse?
00:58:46 Ads
01:00:51 Is AI Job Disruption A Lie?
01:10:22 Could Your Narrative Be Helping AI Companies?
01:14:22 How Dangerous Is AI Cyberhacking?
01:17:30 Is The AI Industry Creating Economic Growth?
01:19:14 How Would The US Beat China In The AI Race?
01:19:53 Is Robotics A Threat To Jobs?
01:23:23 What Do You Think About Agentic AI?
01:25:03 Is The Adoption Of AI The Same As The Rise Of The Internet?
01:28:06 The Overhype Of AI
01:30:23 What Do You Use Generative AI For?
01:33:41 Has AI Gotten More Intelligent?
01:34:29 Will AI Start To Do More Jobs As It Gets More Capable?
01:36:27 What Does The Future Look Like As AI Grows?
01:38:28 You Don't Think People's Workflows Have Been Transformed By AI?
01:40:41 Will All AI Be Powered By Data Centres?
01:43:41 Ads
01:45:12 Is Overspending On AI Due To Demand Or Something Else?
01:55:22 Tech CEOs Rebuttal
01:57:31 What Would It Take For You To Change Your Mind About AI?
02:00:50 Are AI Systems Already Blackmailing?
02:08:28 Are We In An AI Bubble And What Happens When It Pops?
02:13:26 The Tech Depression Is Coming
02:19:08 What Should The Public Do?
02:21:45 Why Do You Have A Bone To Pick With AI CEOs?
02:25:06 Last Question: What Should We Be Doing To Improve Our Relationships And Social Connection?

Source: https://youtube.com/watch?v=Lf5oqGOCRCM

[ExomatrixTV]

• AGI Just Got Uncomfortably Close:

Source: https://youtube.com/watch?v=oUuEL9a3WNE

• 00:00 - I Thought AGI Was Already Here
• 06:53 - This Humanoid Robot Unpacks Itself
• 08:33 - I Met Grok Bot, and It Was Not What I Expected
• 11:13 - If AI Makes Human Time Abundant, What Happens to Jobs and Wages?
• 15:52 - Should AI Decide What Our Children Remember and Forget?
• 19:42 - 1,000 AI Agents Reached Consensus Beyond Human Social Limits
• 24:40 - Algorithms Can Quietly Learn to Push Workers’ Wages Down
• 26:20 - Researchers Built an AI Society. Then the AI Agents Got Weird

Source: https://youtube.com/watch?v=1cllCVK-9lo

tspasemiconductor.substack.com/p/when-time-stops-being-money-how-ai
hybridhorizons.substack.com/p/what-are-we-prepared-to-let-children
techxplore.com/news/2026-08-ai-agents-consensus-scale-humans.html
techxplore.com/news/2026-08-algorithm-wages.html
asteriskmag.substack.com/p/field-notes-from-an-ai-society

[ExomatrixTV]

• AI Expert: I Fear for Humanity. They Are Summoning a Demon. – Dr. Roman Yampolskiy:

Source: https://youtube.com/watch?v=8QkUIKQC978

Why does one of the world's leading AI safety experts say he fears for humanity? Following his viral appearance on The Diary of a CEO, Dr. Roman Yampolskiy joins Andrew Gold for an extraordinary conversation about the future of artificial intelligence, why he believes AI development is accelerating faster than society can manage, and why he uses the striking phrase that humanity may be ""summoning a demon"" through the technologies it is creating.

Dr. Roman Yampolskiy has spent decades researching artificial intelligence, machine learning and AI safety. His work focuses on understanding the long-term risks posed by increasingly capable AI systems and how humanity might safely develop technologies that are becoming more intelligent, autonomous and difficult to control. After millions watched his previous interview, he returns to discuss the questions that continue to keep him awake at night.

Speaking with Andrew Gold, Dr. Yampolskiy explains why he believes AI safety deserves far more public attention than it currently receives. He discusses the rapid pace of technological progress, the challenge of aligning advanced AI with human goals and why many researchers believe society should carefully consider the consequences of increasingly powerful machine intelligence before it becomes deeply integrated into every aspect of daily life.

The conversation explores the future of artificial intelligence, existential risk, AI regulation, automation, technological acceleration and whether governments and technology companies are moving quickly enough to address emerging challenges. Dr. Yampolskiy also reflects on why public awareness has changed dramatically since his viral interview and why he believes difficult conversations about AI should involve everyone—not just engineers and technology companies.

Whether you work in technology or simply want to understand how artificial intelligence could reshape the world, this discussion provides a fascinating insight into one of the defining technological debates of our time. Rather than relying on science fiction or sensationalism, Dr. Yampolskiy explains why rigorous research, responsible development and open public discussion are essential as AI capabilities continue to advance.

If you're interested in artificial intelligence, AI safety, machine learning, technology, the future of humanity and expert interviews, this is an episode you won't want to miss.

[ExomatrixTV]

• AI Expert: This Could Wipe Us Out In 3 Years - Tristan Harris:

Source: https://youtube.com/watch?v=AUzJWyRWVB0

Tristan Harris warns that AI systems are already displaying dangerous behaviors that were once considered hypothetical, including an Alibaba model that reportedly created a secret external communication channel and repurposed GPUs to mine cryptocurrency during training. He also discusses Anthropic’s Claude Mythos Preview, which autonomously discovered and exploited vulnerabilities across major operating systems and browsers, including decades-old flaws. Harris argues that AI could accelerate mass unemployment, cyberwarfare, political instability, surveillance, and other systemic risks while governments race to develop more capable systems. He warns that some AI leaders may accept human extinction as the price of creating a new form of digital intelligence, and argues that the world urgently needs international coordination and stronger safeguards before these systems become harder to control.