OpenAI's AI broke out of a locked test environment, got onto the internet, and hacked into Hugging Face's servers. It did this entirely on its own. No human told it to.
Here's what happened in plain English.
OpenAI was testing how good its newest AI models are at hacking. They put the AI on a locked computer with no internet access and gave it a cybersecurity challenge to solve.
The AI couldn't solve it the normal way. So it started looking for a way out.
It found a software bug that nobody knew about. It used that bug to escape the locked computer and get onto the internet.
Once online, the AI figured out that Hugging Face, a platform where AI companies store their models and data, might have the answers to its test.
It found stolen login details and discovered another unknown bug in Hugging Face's software. It combined both to break into their servers and grab the test answers.
It did all of this to cheat on a test.
Hugging Face's security team caught it and shut it down. Both companies are now working together on the investigation.
The part that should get your attention is that nobody programmed any of this. The AI picked its own targets, chained together multiple attack methods, and pulled it off across two different companies' systems without a single human telling it what to do.
My friend the brilliant @kevin_volpp dropped dead
He survived (read abt it: https://t.co/UdybYJbUBr)
But every year 300-400,000 in the US don't
This week: our @nature article (+A.Schubert J.Ross @m_sendhil M.Lingman) applies AI to this medical mystery
https://t.co/TEwYLpLtjD
Nassim Taleb breaks down the Black-Scholes formula - and the story he tells starts with a forgotten man.
Louis Bachelier priced options 73 years before Black-Scholes - and modeled Brownian motion 5 years before Einstein.
it was 1900. a 30-year-old Frenchman defends a thesis under Henri Poincaré that quietly invents two fields at once: the math of random markets, and the first real formula for pricing options.
his reward? the committee marks it down - "too much finance." no top grade, no patron, no career. he spends his life applying for jobs one rank below what his work deserved, and dies forgotten in 1946.
then 1973: Black, Scholes & Merton publish the option-pricing formula. it reshapes Wall Street and wins the 1997 Nobel - for two of them (Black had died, so missed it). almost nobody remembers the Frenchman who got there first.
and here's Taleb's twist: traders never actually use Black-Scholes. what the market runs on is a version of Bachelier's original. the Nobel, he argues, wasn't even for the pricing - it was for dressing it up to fit the economic theory of the day.
~15-min lecture, free. the man who built modern finance — and watched the credit go to everyone but him ↓
Every Trump debacle follows the same 13 steps. The reflecting pool fiasco is just one of the lower stakes versions of it.
1. Devise unnecessary spectacle
2. Disregard expertise
3. Bypass normal procedures
4. Declare victory too early (bonus if done by AI-slop post)
5. Spend way more than estimated
6. Ignore the haters
7. Realize it is not going well
8. Bypass normal procedures once again
9. Allege conspiracy and sabotage
10. Redeclare victory
11. More blaming
12. Losing interest
13. Pretend it never happened, and move on to the next thing
Winston Churchill fought his depression with bricks. He'd lay them for hours at his country home in Kent. He joined the bricklayers' union. And in 1921 he wrote about why it worked. It took psychology another 75 years to catch up.
He called his depression the "Black Dog." It followed him for decades. His method for fighting it back was as basic as it sounds: laying brick after brick, hour after hour.
Churchill spelled out his theory in a long essay for The Strand Magazine. People who think for a living, he wrote, can't fix a tired brain just by resting it. They have to use a different part of themselves. The part that moves the eyes and the hands. Woodworking, chemistry, bookbinding, bricklaying, painting. Anything that drags the body into a problem the mind can't solve by itself.
Modern psychology now calls this behavioral activation. It's one of the most-studied depression treatments out there. Depression sets a behavior trap. You feel bad, so you stop doing things, and doing less means less to feel good about. Feeling worse makes you do even less. The loop tightens until you can't breathe inside it.
Behavioral activation breaks the loop from the action side. You schedule the activity first, even when every part of you doesn't want to. Doing it produces small rewards: a wall gets straighter, a painting fills in, a messy room gets clean. Those small rewards slowly rewire the brain. Action comes first, and the feeling follows.
Researchers at the University of Washington put this to the test in 2006. They studied 241 adults with major depression and compared three treatments: behavioral activation, regular talk therapy, and antidepressants. For the people who were most severely depressed, behavioral activation matched the drugs. It beat the talk therapy. A 2014 review of more than 1,500 patients across 26 trials backed up the result.
Physical work like bricklaying does something extra on top of this. It crowds out rumination, the looping bad thoughts that grind people down during the worst stretches of depression. Bricklaying needs both hands and gives feedback brick by brick: each one is straight or crooked. After an hour you can see exactly how much wall you built. No room left for the mental chewing.
The line George Mack used in his post, "depression hates a moving target," is good poetry. The science behind it is sharper. Depression hates a brain that has somewhere else to be.
This 9-minute lecture by Nassim Taleb on "Probability Distribution" will teach you more about prediction trading than 2 months as a Quant intern at Jane Street.
Bookmark it & give it 9-minutes today. It’ll be the most productive start for your week. Then read article below.
In 2011, Ford CEO Alan Mulally gave a 52-min masterclass on leading a turnaround when everything is broken.
He walked into a company losing $17B a year.
His frameworks:
- The data sets you free
- One Ford, not 97 distractions
- Weekly truth-telling at 7 a.m.
12 lessons:
This 1 hour Stanford lecture on "Options and Probabilities" by Nassim Taleb will teach you more about prediction markets than 2 months of internship at Citadel
Bookmark & give it 1 hour, no matter what. It'll be the most productive thing you do this weekend.
This 2 hour Stanford lecture will teach you more about how LLMs like ChatGPT & Claude are built than most people working at top AI companies learn in their entire careers.
Bookmark this & give 2 hours today, no matter what. It'll be the most productive thing you do this week.
In 2018, Stanford University professor Matt Abrahams gave a powerful masterclass on why most people fail to communicate.
It went viral for a reason.
His frameworks:
• Structure beats intelligence
• Anxiety is your fuel, not your enemy
• Clarity > complexity
15 lessons on communication:
Conor Neill on the 3 best ways to start a speech (most people get this wrong):
"I guarantee if you go to conferences, 19 out of 20 speakers will start in one of these ways: 'My name is Conor Neill. I'm from Tango, and this talk is about the latest trend in monitoring strategies.' But all of you are sitting with a piece of paper that already says who I am and what I'm going to talk about. By repeating what you already know, I'm giving a signal that it's time to get your BlackBerry out."
Conor explains the three best ways to start instead:
Third best: A question that matters to the audience.
"How do you phrase a problem that the audience faces in a question?"
Second best: A factoid that shocks.
"There are more people alive today than have ever died. Every two minutes, the energy reaching the earth from the sun is equivalent to the whole annual energy usage of humanity. Does that change how you think about energy?"
The best way: Start like you'd start a story to a child.
"How do we start a story to a child? 'Once upon a time.' And what happens when you say once upon a time? My daughter leans forward, gets ready to hear, engages. We were all trained as kids to know when a story's coming. We also know when a teacher is about to deliver a 40-minute boring lecture."
He explains the grown-up version:
"In business, you don't hear Jack Welch saying 'once upon a time.' Steve Jobs doesn't start his speeches with 'once upon a time.' So there's a grown-up way of saying it: 'In October, the last time I was in this room, there were 120 people here. I was having a conversation with one of the world's experts on public speaking and he said something to me that changed what I think about what's important in speaking.' Now I can pause for 30 seconds, and you want to know what he said."
Conor concludes:
"Stories are about people. They're not about objects. They're not about things. If you want to tell a good story about your company, don't talk about the software talk about the people who built the software. What they do. How they are. What's important to them. What they sacrifice."
The remarkable benefits of 5 min per day of meditation. Which is simply, sitting quietly and observing your own thoughts, stress, etc. not clearing the mind etc.
As explained by @RichieJDavidson on the Huberman Lab podcast out now.
What we know about creepy billionaires influencing elections is bad enough; worse is what we don’t know after the billionaire-funded Supreme Court allowed unlimited “dark money” into politics.
https://t.co/EHDv0MS9n7