Accelerating scientific discovery with ChatGPT for Academic Researchers
OpenAI is giving 100,000 academic researchers free access to ChatGPT's most advanced AI models to accelerate scientific research, collaboration, and discovery.
Research
All AI Pulse updates tagged "Research".
41 signalsHow enabling two settings tripled our scores on the ARC-AGI-3 benchmark
How two API settings improved GPT-5.6 performance on ARC-AGI-3, boosting scores and efficiency by retaining reasoning and enabling compaction.
This may be the most consequential period in human history.
This may be the most consequential period in human history. AI is giving scientists and researchers unprecedented capabilities, with progress accelerating rapidly. Worth watching:
Have you spotted us?
Have you spotted us? We're delighted to be featured by @CPOfficialtx as one of the Top 10 Must-Attend AI, Web3 & FinTech Summits in H2 2026. As we celebrate 10 years of World Summit AI, we're bringing together the pioneers, researchers, founders, enterprise le
The most cutting edge research in AI/ML is taking place in industry, not on campuses
The most cutting edge research in AI/ML is taking place in industry, not on campuses
Anthropic Launches $200 Million Fund to Study How Economies Can Adapt to AI Disruption https:...
Anthropic Launches $200 Million Fund to Study How Economies Can Adapt to AI Disruption https:// theeconomicmisfit.com/2026/07/26/ant hropic-launches-200-million-fund-to-study-how-economies-can-adapt-to-ai-disruption/ … Find 80+ verified opportunities with dire
You and Nous Research are truly an inspiration to keep building amazing things I could have n...
You and Nous Research are truly an inspiration to keep building amazing things I could have never thought of before with AI. And, I truly want to keep pushing on this path of open source innovations and make Open Source AI the winner.
How AI is expanding what people do at work
New OpenAI research shows how AI is expanding what workers do, with ChatGPT users taking on tasks across roles and reshaping job boundaries.
On the other hand, it's totally legit for proficient programmers to use AI to help with tedio...
On the other hand, it's totally legit for proficient programmers to use AI to help with tedious work when trying to research, preserve and replicate real hardware. I have no problem with that.
Kevin Clarno will lead a team of eight UT researchers as part of the Prometheus project, a $6...
Kevin Clarno will lead a team of eight UT researchers as part of the Prometheus project, a $60 million effort led by the @INL to reduce the time and cost of developing nuclear energy. Read more:
“2026 年还劝人做公众号”这句背后，真正卡人的常常不是写不出来，是前面 research 太慢。一个主题先拉 10-20 篇同类文章，把标题、首段、结尾 CTA 摊开看，选题判断会快...
“2026 年还劝人做公众号”这句背后，真正卡人的常常不是写不出来，是前面 research 太慢。一个主题先拉 10-20 篇同类文章，把标题、首段、结尾 CTA 摊开看，选题判断会快很多。先把样本池搭起来，再谈批量写，会比一上来让 AI 直接出稿稳得多。
LI is your Ally for life and decades ahead of every AI http:// laminaresearch.com
LI is your Ally for life and decades ahead of every AI http:// laminaresearch.com
CHINA JUST MERGED PAPER AND SCREENS INTO ONE SURFACE.
CHINA JUST MERGED PAPER AND SCREENS INTO ONE SURFACE. AND AI DESIGNED THE MATERIAL ITSELF. Not a tablet that looks like paper. Actual paper that behaves like a screen. You write on it with a pen. It updates like a display. It folds, bends, and feels like the t
SOMEONE BUILT AN OPEN SOURCE STARTER KIT THAT SETS UP YOUR ENTIRE SECOND BRAIN IN OBSIDIAN AU...
SOMEONE BUILT AN OPEN SOURCE STARTER KIT THAT SETS UP YOUR ENTIRE SECOND BRAIN IN OBSIDIAN AUTOMATICALLY no more "where do i even start" you run it, it asks you three questions: → what will you use this vault for? (research, writing, projects, journaling) → wh
A 24-YEAR-OLD STOPPED WRITING AI PROMPTS FOR 30 DAYS.
A 24-YEAR-OLD STOPPED WRITING AI PROMPTS FOR 30 DAYS. HE REPLACED 47 SAVED PROMPTS WITH ONE LOOP THAT NOW RUNS BEFORE HE WAKES UP 6 steps. 3 checks. 1 human approval. His old setup was a folder full of prompts for research, hooks and articles. Every morning he
Open models are no longer just the cheap backup.
Open models are no longer just the cheap backup. Kimi K2 is close enough to frontier coding models that the real question changes: Do you need the best model for this task? Or do you need a strong model that is fast, open, and cheap enough to use every day? Be
This AI research tool @atypica_AI truly helps developers tackle the most critical step before...
This AI research tool @atypica_AI truly helps developers tackle the most critical step before product development—market research. For example, I recently noticed a lot of people working on AI video editing skills, so I did some research on packaging this need
While the US debates export controls on Chinese chips, Alibaba's newest model launched with a...
While the US debates export controls on Chinese chips, Alibaba's newest model launched with a bigger claim than evidence. There's no benchmark table behind that claim. No model card. No license. Just a tweet and a preview endpoint - dropped exactly 2 days afte
Speaking of architecture, I thought of another interesting approach.
Speaking of architecture, I thought of another interesting approach. I directly took a page from Mr. Liang Sicheng's "A Pictorial History of Chinese Architecture" and had the AI create a "magic book." I made the architecture "grow" out of the paper, and labele
anyone can say “built for AI.” the price tells you if it’s true.
anyone can say “built for AI.” the price tells you if it’s true. most search APIs charge $4-$9 per 1,000 calls because they’re running old human search stacks behind an endpoint. Octen charges $1. benchmarks are public, run it yourself
HE STRAPPED 6 AMD MI300Xs INTO A GARAGE RACK AND OUT-RUNS EVERY H100 CLUSTER IN HIS ZIP CODE ...
HE STRAPPED 6 AMD MI300Xs INTO A GARAGE RACK AND OUT-RUNS EVERY H100 CLUSTER IN HIS ZIP CODE ethan is 34, toronto liberty village 2-bedroom above a shawarma spot, ex-vector institute researcher, snagged 6 AMD MI300X modules off a meta resell pallet in the toro
KIMI K3 ISN'T A CHATBOT, IT'S QUIETLY TURNING INTO A ONE-MODEL STUDIO Moonshot AI's new K3 do...
KIMI K3 ISN'T A CHATBOT, IT'S QUIETLY TURNING INTO A ONE-MODEL STUDIO Moonshot AI's new K3 doesn't just answer you, it does the work, and one clip shows the range it spins up fully playable games from a single prompt it turns a deep research query into a worki
Claude Fable 5 vs.
Claude Fable 5 vs. Kimi K3. Both running on the same benchmark via Seedance 2.0. Intentionally pushed them with a 3D voxel environment prompt, but the way they interpreted the complex mechanical layout and internal physics is fascinating. The frontier is movin
Introducing GeneBench-Pro
Introducing GeneBench-Pro, a new benchmark testing AI performance in genomics, biology, and scientific research using complex, real-world datasets.
A PHD STUDENT IN AMSTERDAM TURNED HER OBSIDIAN VAULT INTO A RESEARCH ASSISTANT THAT WORKS WHI...
A PHD STUDENT IN AMSTERDAM TURNED HER OBSIDIAN VAULT INTO A RESEARCH ASSISTANT THAT WORKS WHILE SHE SLEEPS. 3,200 notes. 47 papers in a week. 12 contradictions she did not know were sitting inside her own thinking. It started with Karpathy’s idea of the LLM wi
The Human Creativity Benchmark is now: published on @arxiv dataset on @huggingface accepted a...
The Human Creativity Benchmark is now: published on @arxiv dataset on @huggingface accepted at @icmlconf genAI workshop who else is going to ICML ?
ELON MUSK: If you see any propaganda on 𝕏, tap the Grok icon to fact-check it “You just pres...
ELON MUSK: If you see any propaganda on 𝕏, tap the Grok icon to fact-check it “You just press the Grok icon on any 𝕏 post & it'll analyze for you & research it as much as you want. You can assess whether that post is the truth. Grok is very good at piercing
Anthropic gave an AI a problem their own researchers couldn't crack.
Anthropic gave an AI a problem their own researchers couldn't crack. Humans: 7 days, 23% of the gap. The AI: 5 days, $18,000, 97%. The problem was how to keep a smarter model under control. In other words, controlling superhuman AI. It solved it on its own. Al
In our midyear 2026 economic and market outlook, Qian Wang, Vanguard’s global head of capital...
In our midyear 2026 economic and market outlook, Qian Wang, Vanguard’s global head of capital market research, explains why AI expectations remain high, but concentrated in a narrow set of technology companies. Where could the next wave of gains come from? Wat
How GPT-5 helped immunologist Derya Unutmaz solve a 3-year-old mystery
GPT-5 Pro helped solve a 3-year-old immunology mystery, offering insights into T cell behavior. The breakthrough could support cancer and autoimmune research.
In this episode, @OpenAI Chief Research Officer @markchen90 joins @allenpark to flambé shrimp...
In this episode, @OpenAI Chief Research Officer @markchen90 joins @allenpark to flambé shrimp, cook Korean stew, and chat about being at the frontier of AI research: why scaling laws and pre-training still matter, how OpenAI chooses research bets and allocates
Your analytics tell you what your users did.
Your analytics tell you what your users did. They never tell you why. Today we're launching Poth (backed by @ycombinator ) — the AI-native customer research platform that finds the why.
This diagram shows the paper’s idea that intelligence can be modeled as a small meaning-based...
This diagram shows the paper’s idea that intelligence can be modeled as a small meaning-based decision map, where clues like danger, food, urgency, and location push the system toward 1 of 4 actions: hide, escape, stay still, or move slowly. The point is that
Introducing Computer for Counsel.
Introducing Computer for Counsel. Computer now connects the research databases, document tools, and matter-management systems lawyers use every day. Pull citable sources from @midpageAI , @LegalZoom , @Docusign , @netdocuments , and more. Available for all Pro
Made in London with AWS: Hirsh Pithadia, CEO & Co-founder, @ValyuOfficial .
Made in London with AWS: Hirsh Pithadia, CEO & Co-founder, @ValyuOfficial . Valyu is building a search and retrieval solution for AI and it’s chosen London to do so. Pithadia discusses how the city’s close-knit startup community, research and engineering talen
Anthropic cofounder predicts singularity in 2028.
Anthropic cofounder predicts singularity in 2028. "Claude 10 would build Claude 11 without any researchers involved, only give it the care and feeding of compute and data."
Introducing the Computer Command Panel.
Introducing the Computer Command Panel. Type / to see every mode and skill available in Computer, including Deep Research and Plan Mode. Live now on the web for all Computer users.
As AI takes on longer, higher-stakes tasks, we want models to carry beneficial and safe behav...
As AI takes on longer, higher-stakes tasks, we want models to carry beneficial and safe behavior into new domains beyond their training—and maintain it under pressure. That’s the idea behind our new research on training models to be broadly and persistently be
The Grok plugin for Word is live.
The Grok plugin for Word is live. Draft documents, summarize research from the web and X, or generate diagrams for your writing. Try it now https:// x.ai/grok/word
An OpenAI researcher was mentoring at YC office hours last Tuesday.
An OpenAI researcher was mentoring at YC office hours last Tuesday. I had 8 minutes to pitch. He was the third reviewer. "What's the startup?" "No startup. I'm running a Polymarket bot with Claude Code. Wanted feedback on the architecture." He leaned back. "Sh
In film, "we'll fix it in post" is what you say when something went wrong on set and you don'...
In film, "we'll fix it in post" is what you say when something went wrong on set and you don't want to redo it. AI research has made it our entire methodology: train the model, then patch whatever comes out. Our new ICML oral argues this can't be the basis of