Hongjian Zhou (@itsEmZee_) - ملف تويتر | Zamantika Mersobahis Locabet

تغريدة مثبتة

If this is not obvious enough... > Mar 25, Sakana’s AI Scientist makes Nature > May 13, RSI raises $650M to build self-improving AI > May 13, Adaption ships AutoScientist for model training > May 19, Karpathy joins Anthropic pretraining > May 19, Nature drops 3 AI-scientist papers in one day The next frontier is not bigger models. It’s AI researchers building AI researchers.

English

8

16

203

14.9K

Hongjian Zhou@itsEmZee_·8h

@josephdviviano They must have done somethings right!

English

0

1

25

Joseph Viviano@josephdviviano·13h

@itsEmZee_ I would never have predicted Denmark at #1

English

1

0

1

78

Hongjian Zhou@itsEmZee_·15h

Not surprised, Canadian enterprise AI adoption is still at the “we bought 1k copilots seats but AI can’t really do what we do” stage

Build Canada@build_canada

Canada trails behind many of its OECD peers in business adoption of AI. When our businesses don't operate at their full potential, Canada falls behind on the world stage. We can fix it through policy. 👇

English

9

2

43

7K

Hongjian Zhou@itsEmZee_·9h

> my kid: dad, where were you when AGI arrived? > me: scrolling this old app called X while OpenAI announced it solved an 80-year math problem > kid: did you understand the math? > me: no but I knew the vibes were historic

OpenAI@OpenAI

Today, we share a breakthrough on the planar unit distance problem, a famous open question first posed by Paul Erdős in 1946. For nearly 80 years, mathematicians believed the best possible solutions looked roughly like square grids. An OpenAI model has now disproved that belief, discovering an entirely new family of constructions that performs better. This marks the first time AI has autonomously solved a prominent open problem central to a field of mathematics.

English

0

10

244

9.1K

Hongjian Zhou@itsEmZee_·9h

😂

Nicholas Decker@captgouda24

Math grad student friend comments on the recent Erdős proof.

ART

0

13

1.7K

Hongjian Zhou@itsEmZee_·12h

> May 20, OpenAI solved unit distance problem!

English

1

0

29

1.7K

Hongjian Zhou أُعيد تغريده

Hongjian Zhou@itsEmZee_·20h

If this is not obvious enough... > Mar 25, Sakana’s AI Scientist makes Nature > May 13, RSI raises $650M to build self-improving AI > May 13, Adaption ships AutoScientist for model training > May 19, Karpathy joins Anthropic pretraining > May 19, Nature drops 3 AI-scientist papers in one day The next frontier is not bigger models. It’s AI researchers building AI researchers.

English

8

16

203

14.9K

Hongjian Zhou@itsEmZee_·13h

@_devingannon Super interesting! Good luck!

English

1

0

1

27

Devin@_devingannon·13h

i work in ai enablement & integration within a large enterprise right now. fire up a company. take those learnings and move down market to MM and SMB to help drive adoption. sit with operators that are doing the work. map workflows. educate. integrate. host events. what do you think?

English

1

0

29

Hongjian Zhou@itsEmZee_·13h

Thought would take longer but here it is, breakthrough on the planar unit distance problem.

OpenAI@OpenAI

Today, we share a breakthrough on the planar unit distance problem, a famous open question first posed by Paul Erdős in 1946. For nearly 80 years, mathematicians believed the best possible solutions looked roughly like square grids. An OpenAI model has now disproved that belief, discovering an entirely new family of constructions that performs better. This marks the first time AI has autonomously solved a prominent open problem central to a field of mathematics.

English

0

23

1.4K

Hongjian Zhou@itsEmZee_·13h

@benjamin0chang @Nature @agreeb66 Congrats! Great work!

English

0

1

503

Benjamin Chang@benjamin0chang·18h

My first PhD paper is out now in @Nature! Very grateful to have worked with the FutureHouse team on this, and a big shoutout to my co-first author @agreeb66 😀

English

31

85

934

48.6K

Hongjian Zhou@itsEmZee_·13h

@_devingannon What’s the plan

English

1

0

1

75

Devin@_devingannon·14h

@itsEmZee_ Yeah it’s crazy. I’m working on this

English

1

0

1

74

Hongjian Zhou@itsEmZee_·20h

This is plainly wrong. AI co-scientists will create more jobs in the field. When Lawrence Sperry demonstrated an aircraft could fly straight without hands on the controls, pilots feared for their jobs to be taken. But autopilot did not make pilots obsolete. It made flight scalable. We got more planes, more routes, because the sky was TOO BIG. Same with AI co-scientists. Science is TOO BIG. We don’t need fewer scientists. We need more people exploring more of the unexplored spaces.

Dr. Thomas Ichim@exosome

Today we all lost our jobs..... Three Nature papers showing that scientists in the conventional sense are obsolete At least read the first one.... the AI replaced all things that the scientist does .... nature.com/articles/s4158…

English

3

55

26.3K

Hongjian Zhou@itsEmZee_·16h

@usepapertrace Agreed. don’t think we’re there yet. But optimistic we’ll achieve it.

English

1

0

122

PaperTrace@usepapertrace·1d

@itsEmZee_ What counts as “discovery”? Preregistered hypothesis, novel result, independent replication, and open data and code. Until an agent clears those without human steering, it is a demo, not discovery.

English

1

0

1

36

Hongjian Zhou@itsEmZee_·1d

Probably a year away from AI scientist making autonomous discoveries

Google DeepMind@GoogleDeepMind

We want to help scientists discover their next breakthrough with AI. Gemini for Science is our new suite of experimental tools to help them explore more hypotheses, validate work at scale, unpack literature with ease, and more 🧵

English

1

46

3K

Hongjian Zhou@itsEmZee_·19h

@Jashanx_gill Big apple!

English

1

0

1

49

Jashan@Jashanx_gill·1d

Hi there 👋 If you are new, Let me introduce myself. I’m Jashan, currently based in Ontario 🇨🇦 I post about:  • AI • Building apps • Writing ✍️ I studied Data Analysis for Business 📊 Currently learning about Digital Asset Management 🌁 Also planning to launch a newsletter soon. Still figuring out the exact niche. If you have ideas, let me know 👀 And thank you to everyone who decided to follow me!!

English

37

2

82

8.8K

Hongjian Zhou@itsEmZee_·20h

@TTrimoreau @grok

QAM

1

0

12

Thomas Trimoreau@TTrimoreau·1d

If Claude/ ChatGPT/ Gemini disappeared tomorrow. Who would you actually reach out to for help?

English

186

3

85

7.9K

Hongjian Zhou@itsEmZee_·20h

@letjudycook How is this different from Claude Code but ask it to do design

English

0

422

judy@letjudycook·1d

old enough to remember Claude Design

el.cine@EHuanglu

Claude Design is here powered by Opus 4.7 this is killing thousands of startups..

English

52

23

1.7K

289.3K

Hongjian Zhou أُعيد تغريده

Sungjin Ahn@SungjinAhn_·23h

🧠We introduce "Generative Recursive Reasoning"! Recursive Reasoning Models like HRM, TRM, and Looped Transformers are deterministic — same input, same reasoning, every time. They collapse the entire space of plausible reasoning paths into a single attractor. Our model GRAM (Generative Recursive reAsoning Models) turns recursion itself into a stochastic latent trajectory. Multiple hypotheses, alternative solution strategies, and inference-time scaling not just by depth, but by width — parallel trajectory sampling. And here's the kicker: the same formulation that gives us conditional reasoning p(y|x) also makes GRAM a general generative model p(x). With only 10M params: • Sudoku-Extreme: 97.0% (TRM 87.4%) • ARC-AGI-1: 52.0% • ARC-AGI-2: 11.1% • N-Queens coverage: 90%+ 📄 Paper: arxiv.org/abs/2605.19376 🌐 Project page: ahn-ml.github.io/gram-website w/ Junyeob Baek @JunyeobB (KAIST), Mingyu Jo @pyross0000 (KAIST), Minsu Kim @minsuuukim (KAIST & Mila), Mengye Ren @mengyer (NYU), Yoshua Bengio @Yoshua_Bengio (Mila), Sungjin Ahn @SungjinAhn_ (KAIST)