It's the first time in the history of the world.
For more than 2,000 years, the most intelligent human brain has beenI've been climbing the same mountain and I've been doing math.。
From O'Griggs, Gauss, to Hilbert, to Wales,And every one who conquers a mountain will live a lifetime therein.。
But now,The era of mathematical heroes is over.
XAI co-founder Christian Szegedy's long-running column "Where did math come from?" exploded on X.
In his writings, the mathematician was an explorer of the original volcanic rocks in the jungle.AI is an aircraft that has flown its flag to the top of a mountain that no one has ever climbed.。
The era of lonely climbers is coming to an end, and the era of mapping a whole continent is beginning.
I'm about to graduate in math.It will come out of the Ivory Tower and become the foundation of all scientific and technological advances.
On this day, he took ten.
Take time back to Szegedy's teenage years. At the time, he was taking mathematics classes at the best school in Hungary, eight hours a week, and spent most of his youth on competitive subjects.
Later, he received his Ph.D. in applied mathematics from the University of Bonn.It's been more than a decade on Google.。
Several of the cornerstones in the history of deep learning bear his name.Inception, the classic network for image recognition, was the first author of the paper; he was the first author to reveal that neural networks could be easily deceived by "counter-sampling", and he was the first author of the paper.
In 2023, he becameMusk is one of the 12 founding members of xAI。
But over the course of his career,It's always math.。
In 2019, he wrote a position paper advocating for an automatic formalization of this path.Creating a superhuman mathematician。
What's called automated formalization is that AI automatically translates human-written mathematics into mathematics.Proof of the machine's ability to perform gradual testing。
It's the first time I've seen this.The year 2029。
In 2024, heYou're three years ahead of schedule.Gary Marcus, the AI skeptic, comes back to him below, and I'm not as optimistic as you are.
In February of the same year, he discussed betting with François Chollet, the father of Keras, and others."By June 2026, there will be super-AI mathematicians".。
And then,The year 2026 is here.。
In May, OpenAI overturned the 80-year-old unit distance conjecture. In July, Anthropic researchers used Claude Fable 5.This was the beginning of a new era in the history of the United States.。
By the end of September,Everything started to accelerate.。
On September 4th, the internal model of AnthropicIt took 11 days to write a complete proof of Fermat's theorem in the Lean proof theorem.。
On September 8, OpenAI announced that 10,000 Agents had run for 88 hours.How to solve the Navier-Stokes equation。
On September 21, OpenAI also said that the internal model had already been developed.Solving over 100 longstanding unresolved mathematical problems。
All of this.It's all happening before the year 2029 that he originally predicted.。
Two thousand years, a hero's climbThis is the first time I've ever heard of him.
The first two hours after graduation were the most free time of his life.Then came the emptiness.。
No more competition, no more teaching, no more caring."For the first time, I don't know who I am without it".
He's going to tell mathematicians all over the world,That's what you're going to experience next.。
In the words of Szegedy,Doing pure mathematics is like exploring a desert without a map.It's harder to walk than the Dalan Strait between Panama and Colombia, where newcomers can get lost for days, get malaria, or be eaten by beasts.
The explorers did three things:Drawing maps, mining, making tools。
The most revered is to climb a high mountain as hard as possible.The man who sketched on top of the mountain。
And the mountains which were lost before them.Whoever climbs it will be honoured.This is what the Langlands program accomplishes.
To find "minerals" that are useful to ordinary people, such as finding elliptical curves that can be used in cryptography, most mathematicians don't care much.I'll just put it in the application for funding.。
The stones that people want to conquer are guesses. They are often not so interesting in themselves.It's a testing ground for rock tools.。
But the explorers stayed in the jungle, and the picture was something else.
After all, this is why people are left behind. Not for the sake of a map, not for the sake of money, but for the sake of that difficult moment, when an unforeseen stone becomes a great mountain, and the whole landscape unfolds under your feet.
In the words of Stephen Wolfram, the father of Mathematica,Pure mathematics is the largest intellectual dam ever built by human civilization.。
It was a lonely climber for more than two thousand years.It's a rock, it's a hill, it's the foundation.。
A plane raises its flag to the top of a mountain
And then,The plane's here.。
The Wright brothers' first flight was only 12 seconds away from the wingspan of a large passenger plane, Segedy said.That's pretty much where AI is in mathematics today.。
Until this year,These planes are still raw.It can only fly a few hundred meters, a few meters from the ground, is incredibly expensive and can only be used on a few occasions.
But over the past few months, they have been able to fly into the canyon without crashing, "we have even landed on a few rocks that no one has ever climbed".That's how I've solved a few myths.。
Then he wrote a parable.
I tell you the truth, these stones are not so interesting, but we climbed them to give you a real tool to grind mountains, and your planes could not fly that high.
Szegedy's judgment was that "airplanes can't have a real mountain top, and they can't draw a big map".It is a fallacy of static thinking.Today's planes are more primitive than they were a few months ago, let alone a few years later.
Graduated in math.
When the plane comes down cheap, it's not easy.The whole jungle will change.。
Szegedy predicts that private jets everywhere, and anyone who can afford to pay, can land on any rock, and humans will invent satellites.Map the entire forest's mineral deposits and then mine them.。
Today's explorers will be frightened and miss the good old days.It's a big company that's ruining the jungles they love.。
He wrote, "They are right, but that doesn't change anything.Big companies will find unimaginable wealth there。」
The rocky movement will not go awayThere will be more amateur explorers flying to their favorite stones, buying ready-made equipment from tool shops, and climbing with friends.
People who are a little physically weak get on planes looking for bigger mountains, orFrom a place higher than any mountain, overlooking an entire mountain range。
"The atmosphere and excitement of the old days will be gone forever. Mathematics will be better than the old days of heroic adventure".It's more fun, more useful, and more ordinary.。」
The first sentence of the last paragraph of the sentence is:"Don't mess with planes".
Those who say AI doesn't speculate, doesn't build theories, doesn't explain, describe models from last spring, and within months, those descriptions don't hold up.Whoever puts his identity on "AI can't do X" is building a house in the sand.
University, journals, lifelong learning committees, in a few years' time, they'll look pretty much the same.What they reward will change fundamentally.。
This kind of heroism becomes less important when you climb mountains that no one else can climb.Choose which terrain is worth exploring, command a squadron, and tell the world what you've found.。
He also acknowledged that "for many people in the system,It would feel like a downgrade.Some people will fight".
The last paragraph of the sentence is as follows:
"Mathematics is about to graduate. It will grow into a full-fledged industry.To become the real infrastructure for all scientific and technological progressIt has become the foundation upon which all other fields stand, and is no longer a quiet corner of the academic community.
The era of the lone climber is coming to an end, and the era of mapping the entire continent is beginning.I can't wait to get on a supersonic plane to see the vast mountains.。」
Stanford mathematician Jared Duker Lichtman, whom he named after, wrote in a transcript that he had been a graduate student since 2019.I'm waiting for this moment.。
Only subversion, without replacement, will leave a void. But a clear vision of a future that will bring new life, the next step is already clear: we are ready to transform all known mathematics into formalized code.
We have to know, we will eventually know.
However, not everyone is willing to receive this diploma.。
Just five days before Szegedy's speech, Wolfram wrote a lengthy statement of more than 7,000 words, saying that he was saying that "there is no need for people to do math anymore"."I'm getting a little impatient".。
In his view, the goal of mathematics "almost by definition has to come from outside the system, specifically from us".The evidence is not what he wants to prove.。
He concludes:Pure mathematics will never be done.。
The Imperialist mathematician, Kevin Buzzard, used"Fifth Stage of Sadness"For the entire mathematical community, the first four stages are represented by denial, anger, bargaining, and depression.
He noted that Fields Prize winner Scherzer publicly stated that he would not use AI, "willing to die for it".
Another doctoral student, whose research card proved it in one argument, ChatGPT once. The doctoral student began to doubt the meaning of what he had done.
This is the sadness that Segedy knows better than anyone.The day he graduated from high school, he also thought he had lost a part of his life.
But he said in his obituary that the grief was real.It's just not going to last long.。
In 1930, Hilbert left a note in Königsberg, which he later engraved on his tombstone."We've got to know, we'll finally know".
Buzzard also quotes the phrase at the end of the sentence.
Mathematics is destined to be the first basic discipline to be industrialized by AIIt can be tested by machines step by step, without experimentation, right is right, wrong is wrong.
So the first hard bones of AI, the millennium puzzle, the Jacobite conjecture, the Fermat theorem, are all in mathematics.
Mathematics is moving from a handful of geniuses to the underpinnings of the entire tech industry. It was the first continent to be mapped on the road to superintelligence.
RoboInd editorial channelResearch, robotics companies and market intelligence for cross-border sourcing.Explore the channel ↗
Share this
End of editorialExplore more from Global Robotics News ↗