IN 1999 MIT FILMED A MATH LECTURE THAT QUIETLY BECAME THE FOUNDATION OF EVERY AI MODEL YOU'VE EVER USED AND ALMOST NO ONE WAS TAUGHT TO SEE IT THAT WAY
39 minutes from Gilbert Strang, who taught this at MIT for over 60 years -- the linear algebra course an entire generation of engineers and data scientists grew up on.
-> The shift it creates: you stop seeing matrices as boring grids of numbers and start seeing them as the language of space, data, and motion itself.
School drilled you to crunch matrices by hand and never told you why. Strang shows you what they actually mean.
Every neural net, every embedding, every model you prompt is linear algebra running underneath. The math you skipped is the engine of the thing you use all day.
Memorizing the steps was never the skill -> seeing what the numbers do is. This is where it finally clicks.
Most people fear linear algebra and move on. The ones who watched this see straight into how AI actually works.
Bookmark & Watch it today, this one's a legend ↓
Näkymätön jarru taisi juuri poistua.
Tilastopäivitys korjasi vuosien 2024-2025 BKT:n selkeästi plussalle.
Kaikki viime aikoina julkaistut ennusteet (ml. VM 16.6) perustuvat liian ankeisiin lukuihin historiasta.
Odotettavissa lähiaikoina: kilvan paranevia talousennusteita.
The smartest man in AI just exposed the whole AGI narrative as a LIE.
And he used a physics problem from 1905 to prove it.
His name is Demis Hassabis. He runs Google DeepMind, and won the Nobel Prize for using AI to crack a problem in biology that had stumped scientists for 50 years.
Almost nobody in this industry has a track record like his.
He went on the NothingButTech podcast and called out the biggest lie in AI right now:
Right now the loudest voices in AI are telling you that AGI is basically here. OpenAI has literally defined AGI as a system that can outperform humans at most "economically valuable work." In other words, if it replaces enough jobs, we have arrived.
Hassabis thinks that bar is a joke.
He said real general intelligence has to do what the human brain can do, because the brain is the only proof we have that this kind of intelligence is even possible. He called that "a higher bar than just being able to do some useful economic work," which is about as close as a polite British Nobel laureate gets to calling his rivals out.
Then he gave the actual test:
Today's AI has read everything humans have ever written, including the theory of relativity. So when it explains relativity back to you, it's repeating an answer that already exists.
That's not intelligence.
So Hassabis proposed a test that makes memorization impossible. Train an AI on only what humanity knew in 1901, four years BEFORE Einstein published relativity. Then ask it to come up with relativity on its own.
It can't look up the answer, because in 1901 the answer doesn't exist yet. The only way to pass is to do what Einstein actually did: Take the same physics everyone else had and reason its way to an idea no human had ever had.
Hassabis says not a single AI today can, no matter how much it has memorized. Which means what we keep calling "almost AGI" is really just the best librarian in history.
It can find any answer that already exists but it cannot create one that doesn't.
His second version is even sharper:
AlphaGo, the system his own team built, famously invented a brand new move that no human had played in 2,000 years of the game.
Everyone called it genius but Hassabis says that still is not the bar.
The real test is not whether an AI can invent a new move inside Go, it is whether an AI could INVENT a game as deep and as beautiful as Go in the first place.
No model that exists today can do it.
The people telling you AGI has already arrived are the same people raising hundreds of billions of dollars on that exact promise.
The valuations only work if the finish line is right in front of us. So the finish line keeps getting dragged closer, and AGI keeps getting quietly redefined down to "does useful work," until the products they already sell happen to qualify.
Hassabis has nothing to prove and nothing to sell you. He already won the Nobel, and he is telling you the machines still cannot do the one thing that would make them genuinely intelligent, which is have a truly original idea.
To be fair to him, he is not a pessimist about it. He believes real AGI IS coming, and he is spending his life building it. He just refuses to pretend it is already sitting in your phone.
So the next time a founder tells you AGI is months away, remember that the one man in the room with a Nobel Prize built his test around Einstein, and admitted that nothing we have made can pass it.
What do you think?
A freelance journalist who had never taken a statistics course wrote a 142-page book in 1954 that professional statisticians still hand to students before anything else, because nobody before him had bothered to explain the tricks in plain language.
His name was Darrell Huff. The book is called How to Lie with Statistics.
I read it in one sitting and spent the next three days noticing the tricks everywhere.
Over 1.5 million copies have sold in English alone. It became a standard college textbook in the 1960s and 70s. Seventy years later it is still in print, still assigned, still the first thing a working statistician reaches for when they want to teach someone to think clearly about numbers.
The man who wrote it was not a researcher. He was a freelancer who wrote how-to articles for magazines. He had no PhD, no academic post, no institutional affiliation. He just understood that numbers could lie without technically being wrong, and he thought someone should explain how.
His opening line sets the whole tone of the book.
"The crooks already know these tricks; honest men must learn them in self-defense."
That one sentence is the entire argument. The manipulation is not coming. It already happened. It happened this morning in the article you read and the chart someone showed you at work and the study your doctor quoted. The only question is whether you know what to look for.
Huff called the first trick the Well-Chosen Average.
When someone tells you the average salary at a company is $80,000, they have told you almost nothing. If the CEO earns $2 million and the 20 employees earn $30,000 each, the mean is $80,000. The median is $30,000. Both are technically correct. One is a lie. The person reporting the number chose which average to use, and they almost always chose the one that served their argument. Huff's rule: whenever you see an average with no description of which average it is, ask.
The second trick he named the Gee-Whiz Graph.
A line chart shows company profits rising. The line shoots nearly vertical, almost doubling in height across the chart. You feel impressed. Then you look at the y-axis and notice the chart does not start at zero. It starts at 94. The actual increase in profits was 3 percent. The dramatic visual was produced entirely by cropping the bottom of the chart. Nothing in the data changed. The picture changed everything.
Every news organization on earth still does this every day.
The third trick is the one that should change how you read every study you ever encounter. Huff called it Post Hoc Rides Again, which is short for the Latin phrase post hoc ergo propter hoc. After this, therefore because of this.
Cities with more churches have more violent crime. Therefore churches cause violence. The logic is airtight. The conclusion is absurd. Both church attendance and crime go up as population grows. The two numbers track each other because a third variable drives both. The correlation is real. The cause is invented.
Huff showed that this structure is not a rare mistake. It is the default pattern of almost every study reported in a newspaper, because causation is a boring word and because proves is a better headline than correlates with.
The fourth trick was the one that floored me. He called it the Semi-Attached Figure.
A headache pill company claims their product is twice as fast as the competition. The study behind the claim is real. The product was tested and the numbers are accurate. What the advertisement does not mention is that the study measured absorption rate into the bloodstream, not relief of headaches. The two things are related but not identical. The statistic is real. It is attached to the wrong conclusion.
Huff said this is the most dangerous trick of all because the number is never fabricated. You cannot fact-check a semi-attached figure by verifying the statistic. You have to ask whether the statistic actually measures what the claim requires it to measure.
Almost nobody asks.
There is one part of Huff's story that most people who recommend the book leave out.
Years after he wrote it, he was hired by the tobacco industry. He worked on a follow-up manuscript called How to Lie with Smoking Statistics, designed to cast doubt on the research connecting cigarettes to cancer. The book was never published. He testified before Congress in an attempt to undermine the statistical evidence against tobacco.
The man who wrote the clearest guide to spotting statistical deception spent the end of his career deploying those same tricks against evidence that was killing people.
That detail does not make the book wrong. The tricks he described are real and the defenses he taught are still the right ones. But it is a reminder that the tools in the book are neutral. Understanding how lies are built does not protect you from choosing to build one.
The crooks already know these tricks.
Some of them wrote the manual.
What is one statistic you have seen recently that you now think deserves a second look?
🚨 SCIENTISTS EXPLAIN: Light doesn’t actually “slow down” in glass time does. And that’s exactly why rainbows exist.
For centuries we were taught that light slows down when it enters glass or water, causing refraction. But the deeper reality is more beautiful: light still travels at c between atoms. What changes is the time delay caused by constant absorption and re-emission by the material’s electrons.
This tiny delay is different for every wavelength → which is why white light splits into a rainbow.
Why this matters:
• In vacuum, light always travels at c
• In glass/water, the phase velocity and group velocity appear slower due to interactions with matter
• Different colors (wavelengths) experience different delays → dispersion
• This is what creates rainbows, prisms, and the beautiful colors we see in nature
The deeper implication is mind-bending:
Light doesn’t “slow down” like a car hitting traffic. It’s constantly being absorbed and re-emitted by atoms, and the accumulated time delay reshapes how the wave propagates. The universe uses time itself as a tool to bend light and paint rainbows across the sky.
What do you think is this one of the most elegant explanations in physics?
Follow for more frontier optics and quantum explanations.
THIS IS WILD!🤯
A solo founder built a 7-person company inside Claude Code.
AI agents find leads, write emails, create content, run campaigns, and monitor operations.
One person. An entire company.
A YouTuber with 110 million subscribers released a free version of ChatGPT.
His name is Felix Kjellberg. You know him as PewDiePie.
He spent his own money on a 10-GPU computer at home. He used it to run the same kind of AI models that power ChatGPT, but on his own hardware. Then he wrote his own app to chat with them, because the apps that already exist were not good enough.
Then he gave it away for free. Anyone can download it. Anyone can change it. Anyone can run it.
It's called Odysseus.
It runs on your computer. Your data stays on your disk. No account. No tracking. No monthly fee.
What you get:
- A chat window like ChatGPT
- An AI assistant that can browse the web, read your files, and do tasks for you
- A tool that scans your computer and tells you which AI models will work on it
- A research mode that reads many websites and writes you a report
- A side-by-side mode to test two AI models on the same question
- A writing editor where AI helps you, instead of writing for you
- Memory, so the AI remembers your past chats
- Email with AI that sorts your inbox and writes replies for you
- Notes, a to-do list, and a calendar
- Works on your phone too
23,612 stars on GitHub in 2 days. Top of trending all weekend.
ChatGPT Plus costs $20 a month. Claude Pro costs $20 a month. PewDiePie's version costs nothing, runs on your own computer, and the code is open for anyone to read.
This is what AI looked like before the subscription model.
(Link in the comments)