Mathematician Terence Tao:
"we have to slow down AI. the pace is insane, and there's no reason to be this fast — no reason at all"
It's amazing how willing we are to change everything without any idea what happens afterward
These are extremely nonlinear dynamics
"we have to slow down AI. the pace is insane, and there's no reason to be this fast — no reason at all"
It's amazing how willing we are to change everything without any idea what happens afterward
These are extremely nonlinear dynamics
Sep 19, 2026 · 5:50 PM UTC
882
1,029
7,430
7,578,293
@haider1 "We should ban the printing press" - a scribe
8
18
1,161
18,893
@haider1 The answer isn't slowing down AI, it's collapsing the fake dreams of a few losers running US AI labs who are willing to salt the mathematical earth if it gives their IPOs a 2% higher chance of succeeding. Pop the bubble and this fixes itself.
21
43
663
40,592
@haider1 Its cause the geniuses dont feel special anymore, AI is coming for everything.
8
14
1,192
23,458
@haider1 "We have to slow down the machines. The pace is insane, and there's no reason to be this fast — no reason at all."
5
5
735
12,615
@haider1 There is a reason. We need very fast progress in medicine. Many lives are at stake. Not possible without ASI.
19
8
723
22,466
@haider1 This is the first time in history that scientists are not the sole drivers of progress and discovery, and -coincidentally, for sure...- it's the first time they have argued that we should slow down.
I had never seen a researcher even think this before last month.
I had never seen a researcher even think this before last month.
18
17
386
19,592
@haider1 There is an excellent reason for the pace to be this fast, including the fact that there are about 1200 people dying of cancer every hour. I have heard people say “no, no, we can cure cancer without AI.” Well, they are welcome to demonstrate that. They will not be able to.
4
7
104
3,702
@haider1s no reason to be this fast."
amazes me the way that people who are very smart in other fields simply cannot comprehend markets and the forces that drive them.
AI needs to go this fast for the same reason you can ship a kiwi fruit half way around the world and still sell it for 30 cents and, similarly, cannot and should not be stopped for the same reason.
amazes me the way that people who are very smart in other fields simply cannot comprehend markets and the forces that drive them.
AI needs to go this fast for the same reason you can ship a kiwi fruit half way around the world and still sell it for 30 cents and, similarly, cannot and should not be stopped for the same reason.
2
0
104
2,742
@haider1 Slow down on AI for math? Lol…it’s not like math has any *direct* influence on anything in the real world. What levers are going to get pulled that increases risk?
Fear-mongering nonsense.
It’s career insecurity. Nothing more.
Fear-mongering nonsense.
It’s career insecurity. Nothing more.
7
3
82
6,636
@haider1 hundreds of millions dying of aging and illness every year disagree that there is no reason not to slow down.
5
2
272
6,909
@haider1 most mathematicians don't understand how humans work
1
0
13
828
@haider1 I guess curing desease, extending life and automating the current miserable jobs that people have to do don't count huh?
Must be nice to be that comfortable!
Must be nice to be that comfortable!
1
0
9
829
@haider1 There are still way too many bottlenecks in the world that are basically intelligence problems.
I’d rather keep pushing capability and get much better at containing the genuinely risky parts than slow the whole thing down.
I’d rather keep pushing capability and get much better at containing the genuinely risky parts than slow the whole thing down.
0
0
4
609
@haider1 It's not every day you get to recognize a brilliant scientist as a complete dumb ass human being.
1
0
4
392
@haider1agree but i totally get him
i got my first computer back in 90s when I was 5, and i was so addicted to programming at that time. I made numerous apps using Visual Basic and c# as a kid and others looked at me like an alien.
but right now there is no way to get this kind of intellectual satisfaction. what i did, every kid can do it now.
i got my first computer back in 90s when I was 5, and i was so addicted to programming at that time. I made numerous apps using Visual Basic and c# as a kid and others looked at me like an alien.
but right now there is no way to get this kind of intellectual satisfaction. what i did, every kid can do it now.
0
0
3
724
@haider1 笑死 数学界最激进的eacc 一个星期前都在allin ai现在自己的职业被一锅端了 开始变减速投资了 真是嘻嘻哈哈
0
0
1
196
This is a wildly out-of-context quote used to sensationalize Tao's otherwise constructive criticism of frontier labs' approach to AI Math.
The quote came from the Q&A session of a talk where Tao discusses the challenges of working with AI in "Pure Mathematics". Tao noted that:
- "Pursity-driven research.
- Solving open problems in pure math typically does not lead to immediate applications, and therefore is not in and of itself interesting. **
- Recent results, like the solution to Navier-Stokes, are missing the important "proof digestion" step (see graphic).
So what Tao is actually saying is: for "Pure Mathematics", there is no need to further speed up "proof generation" and "proof verification", without first scaling "proof digestion".
** This is different than most sciences, where solutions to problems can be very impactful, like curing cancer for example!
My views:
- I'm an accelerationist, but I agree with Tao here, since pure math has no direct applications.
- There are many open problems in applied math (e.g. theoretical computer science & cryptography) that do have more direct applications (proximity gap for Reed-Solomon, for example). We should accelerate AI problem-solving on problems with downstream impact.
0
0
2
397
@haider1 He greater the raw iq the greater the cope
> most PhDs are a waste of time at this point
> if you value intelligence above all other human qualities, you’re gonna have a bad time - ilya
what's crazy to me to witness now (though not surprised), is that objective math and soon, physics & science are cooked. Now that objective problems are easy to define for AI to hill climb + bulk solve, even hardest problems humanity couldn't before
BUT the messy worlds in which problems are not so easy to objectively define, software products included, are NOT cooked. Much steering left with ai + humans BEFORE there's agi if there ever is one: youtu.be/AxzcWOxzkiw?si…
0
0
1
323
@haider1icians are about to become more valuable, not less.
Not because they'll out-prove AI. Because someone has to stand between machine output and human decisions and say precisely what was shown, under which assumptions, and what wasn't.
Interpretation becomes the scarce job.
Not because they'll out-prove AI. Because someone has to stand between machine output and human decisions and say precisely what was shown, under which assumptions, and what wasn't.
Interpretation becomes the scarce job.
0
0
0
73
@haider1 You can’t see ANY reason to be this fast?
1
0
192
4,909
@haider1 Even the best mathematician can be stupid outside his area of expertise. Don't be an expert, it makes it hard to think...
3
6
170
7,879
