rx math
Region: US
Monday 20 July 2026 17:43:59 GMT
44297
2368
86
186
Music
Download
Comments
chrismcdaniel457 :
wait your saying ai disproved the jacobian conjecture!? wow!!!
2026-07-20 21:08:32
41
Illidamn :
the exact type of thing AI should be used for
2026-07-20 19:44:21
176
kg :
this is where AI shines the most: hard to solve but easy to verify a solution
2026-07-20 20:19:06
78
Max :
Supposedly this is how it came to its conclusion: Take π: P¹ × Sym²(P¹) → Sym³(P¹), (p, {q,r}) ↦ {p,q,r}. R be its ramification divisor; H ⊂ Sym³(P¹) ≅ P³ be hyperplane tangent but not osculating to the small diagonal; X := (P¹ × Sym²(P¹)) \ (R ∪ π⁻¹(H)) ≅ A³; Y := Sym³(P¹) \ H ≅ A³. π|X: X → Y is counterexample
2026-07-21 00:26:42
5
dakine.surfer :
I wonder if these discoveries are gonna accelerate…
2026-07-20 21:59:43
5
imaginaryman :
When I was five, my grandmother used to sing a song to be about an existence proof that the Jacobian conjecture was false. I don’t remember the words, and you sing me to sleep? 🥰
2026-07-21 04:08:42
28
Jake :
It’ll be interesting to see Anthropic and OpenAI trying to one up each other with math research. I think this is the first major one out of anthropic
2026-07-20 20:46:52
5
3y3less :
Obviously this counterexample was included in some paper somewhere forgotten in some old paper maybe about other topic. It just seen it before... (I'm building LLMs and different AI models proffessionally, I'm 99% sure that this is how it worked this time)
2026-07-21 06:25:05
0
Timothy Mattnew :
So now we are robbed of a beautiful constructive process that might have discovered new theories along the way, and we just get the final result and verify it.
2026-07-22 08:17:25
1
Cedric :
It definitely has seen a lot of polynomials and associated jacobians, which will substantially cut down on number of tests it'll need to produce.
2026-07-20 20:52:36
4
M. :
I would question it because of the rounding errors in computational math.
2026-07-20 21:13:18
0
AAA :
What is this used in real life?
2026-07-21 12:01:59
0
Thibaut :
AI will soon advance science much further and much faster than humans can. Next year, we'll have the manual for generating antigravity, traveling in time or being immortal.
2026-07-20 18:30:45
5
Baelor I Targaryen :
What does this mean for space travel
2026-07-25 02:05:28
1
hyejin hadid :
Maybe they found it themselves and credited Claude to promote it
2026-07-21 00:00:47
0
Lippspa :
I'm not even in stem smh
2026-07-20 18:04:42
2
berrysandcream69420 :
Some guy probably solved it 20 years ago on Reddit and it was in the training data 😂😂
2026-07-23 22:57:08
7
xxx :
i checked and it works, you can replicated it quite easily
2026-07-21 17:01:11
0
a456080 :
In several fields, gen AI has been incredible specifically at constructing examples and counterexamples. I find it still quite weak for trying to write new full proofs
2026-07-21 00:13:11
1
azaral1 :
it's not solved there is a difference between observable and state space
2026-07-29 08:54:54
0
ghostinthefeed :
i love how ge van came up with something and thendont want to tell real source
2026-07-21 07:33:17
0
anto :
This is crazy
2026-07-21 01:05:20
0
SciencePen :
11 years ago, I dreamed of a powerful search tool that could explore the entire mathematical literature and identify techniques that might be transferred from one field to another. Most mathematicians cannot follow every development outside their own area, yet a method from one branch of mathematics may turn out to be extremely useful in another. Some important papers are also difficult to read because they use unfamiliar notation, so I imagined a tool that could translate their ideas into notation familiar to the reader. AI is trained on vast amounts of mathematical writing, which gives it access to ideas, methods, and connections across many different fields. It is becoming the tool I was dreaming about: one that can help researchers search across disciplines, explain difficult papers, translate notation, and even generate counterexamples. This last ability could be especially valuable, allowing mathematicians to test their conjectures before investing significant time in trying to prove them.
2026-07-21 12:29:04
0
yyuu779 :
I thik it’s brute force , he got unlimited tokens
2026-07-27 20:32:20
0
To see more videos from user @rx_math, please go to the Tikwm
homepage.
