Roy @roydherbert

Future Fields Medallist and Millennium Prize Winner, but mostly a Time Lord who lost his screwdriver. Creator of the Continuity Model, aka Chronoflux. Earth. Joined February 2022

Tweets

72K
Followers

10K
Following

294
Likes

153K

Freddy Vega @freddier

12 hours ago

GPT-6 Astra got +80% on a Physics benchmark. It insisted the expert results were wrong and its calculations were right. A team checked. And yes. GPT-6 Astra was right. New score: 98%

Pratik Rath @PratikRath15

a day ago

arxiv.org/abs/2610.16733

23 103 1K 804K 556

109 665 11K 540K 2K

View Details

Newton didn't propose a theory of gravity, neither has Einstein, in fact it remains problematic for mush of Gravimetry research. I am only aware of one model that does propose a resolution for gravity which is three decades old, but it it not widely known. With my regards - todays Newtonian representative.

0 1 0 34 0

View Details

Roy @roydherbert

8 hours ago

François, with respect that is very close, but I would put the failure one level deeper. The problem is not simply that intelligence lacks sensory input. It is that most formal systems have no internal requirement that what they construct remain continuously recoverable from the same physical parent. That is precisely the problem I set out to solve. In Chronoflux, admissibility constrains what may exist, recoverability demands that the structure can be obtained from the underlying physical dynamics, and persistence determines what survives sufficiently to become a measurable record. As such, grounding is not an extra stream of data fed into intelligence after the fact. It is built into the architecture of what may count as knowledge in the first place. An intelligent system can then speculate as freely as it likes, but nature retains the veto François. x.com/fchollet/statu…

François Chollet @fchollet

17 hours ago

String theory is what happens when an intelligent system is starved of sensory input and starts hallucinating. Intelligence needs grounding. For this reason, intelligence is practically never the bottleneck to solving impactful problems. The bottleneck is iterated contact with

662 1K 10K 4.4M 3K

0 2 3 491 0

View Details

Riccardo De Santi @r_de_santi

13 hours ago

Joint work with amazing collaborators and advisors: @fededg__*, @gna_ngelo*, @_sophia_tang_, @pranamanam, @FannyYangETH, @ismaililkanc, @mmbronstein, @arkrause. *shared first authors, alphabetical order. (6/6) Paper: arxiv.org/abs/2610.17246 AlphaXiv: alphaxiv.org/abs/2610.17246 Website: llm-to-flow-distillation.github.io Code: github.com/FedericoDiGenn…

0 3 24 1K 8

View Details

Simone Conradi @S_Conradi

11 hours ago

I’ve never been able to draw, until I learned to turn math ideas into code and explore visualizing things. More than 20 years ago I started with @WolframResearch Mathematica, now ~1000 #Python scripts: from complex analysis to statistical mechanics to fractals. #builtWithMarimo

6 21 251 4K 43

View Details

Aaron Kesseler @arrowbombadil

a day ago

My Grok @bot made this video 🎬🎙✂️ And my other one has its own crypto wallet now 🪙📈🤖👀 From @Hightv's HIGH ON GROK BOT🚨 𝕏 Space. ✅ Soon you can make your own clips from any 𝕏 Space 🎙

3 5 11 597 4

View Details

Roy @roydherbert

16 hours ago

@DIYDisclosure As You Like It, Act II, Scene VII 😉

0 0 3 56 0

View Details

Roy @roydherbert

16 hours ago

@DIYDisclosure Consciousness is it's audience, yet we are all our own director and play the leading role in it's theater.

1 1 6 140 2

View Details

Roy @roydherbert

16 hours ago

One thing leads to another, yet many are still discussing the phenomena, and not what it informs us. That should beg questions, never mind the what but firstly, how. If you can explain the how, then that opens up possibilities and so on. Then you're faced with the science, after that comes the bigger picture and so on.

1 1 3 62 1

View Details

Roy @roydherbert

17 hours ago

@LeonidasPlatan1 That would be a burden, better to hold continuity alone, infinity naturally follows.

1 0 5 107 1

View Details

Polymarket @Polymarket

2 days ago

JUST IN: Fields Medal-winning mathematician reveals he cried after OpenAI released hundreds of proposed solutions to unsolved math problems, declaring, "They have annihilated my field of research."

512 674 9K 530K 931

View Details

Jens Eisert @jenseisert

2 days ago

One of my favourite courses to teach is elementary quantum mechanics. It offers students their first encounter with a physical theory of nature—arguably the best we have—that makes extraordinarily precise predictions, yet defies our everyday intuition. It has been a joy to teach this course in a packed lecture theatre. And it was for this important course that I had to leave the @FU_Berlin retreat. physik.fu-berlin.de/en/einrichtung…

3 3 66 3K 5

View Details

Roy @roydherbert

2 days ago

@hahabirdpumpkin Hopefully citing my work, but alas human being's are what they are.

0 0 2 71 0

View Details

Roy @roydherbert

2 days ago

@zennospace Therefore one will require an understanding of temporal calculus. Mechanics of Time: The Full Recovery of Nature (Chronoflux) amzn.eu/d/059SB37D #Amazon via @amazon

0 1 5 87 0

View Details

Zenno Astronautics @zennospace

3 days ago

In this house we believe space-based superconducting magnets are a time traveling god that is actively reaching backwards through time to construct the conditions of its own Kardashev Ascension.

34 45 435 16K 60

View Details

Roy @roydherbert

3 days ago

With it still trending I should perhaps remind the world of this one. @ClayInstitute @OpenAI zenodo.org/records/213565…

1 1 10 487 5

View Details

Roy @roydherbert

3 days ago

@kyronis_talks It doesn't end there my friend. We welcome these results as further confirmation. zenodo.org/records/213475…

0 0 2 82 1

View Details

Michael Bronstein @mmbronstein

3 days ago

@ziv_ravid You cannot undo human stupidity with artificial intelligence

4 2 25 2K 1

View Details

Roy @roydherbert

3 days ago

Quite, my gifted friend, intelligence can extend capability, but it cannot supply judgement where none is being exercised. The more powerful the machine becomes, the more important the quality of the human question, assumption and intent becomes. As such, artificial intelligence may amplify intelligence, it will amplify stupidity just as efficiently.