Every week, I organize 400 to 700 links into roughly 60 categories as part of my ongoing effort to learn about AI. This is my personal notebook, which I enjoy sharing with friends… a hobby and a labor of love, rather than a commercial publication or product.
If you arrived here through a search or shared link, this page collects the links I found for AGI for the week ending July 31, 2026.
As part of my learning process, I like to automate the category covers. It gives me a chance to learn Python and APIs.
This week’s cover prompt was written using Claude Opus 4.7, and the image was generated using Gemini 3.1 Flash Image Preview.
Category Cover Image Prompt: Long-lens anamorphic night shot down an empty elevated Los Angeles freeway that terminates at a black horizon, a 1973 Chevrolet Chevelle Malibu in matte primer gray idling alone at the vanishing point with glowing taillights, city grid shimmering below in sodium amber and deep teal, a faint golden-yellow scorpion-shaped constellation above, hot magenta handwritten brush-script title 'AGI' scrawled in the upper third, 35mm grain, soft halation, crushed blacks, cinematic 1980s neo-noir poster composition with generous negative space.
This Week in AGI News
Summary by Claude Sonnet 5.5 based on this week’s links:
- Models improving their own tools: Cline said Kimi K3 spent 17 hours rewriting the Cline harness, lifting its Terminal Bench score from 77.5% to 88.8% and cutting run cost from $79 to $49.8. OpenAI said GPT-5.6 Sol made itself cheaper to run, with 20% lower serving costs from GPU kernel work.
- Labs ask for a way to pace the frontier: Anthropic and OpenAI both backed a petition signed by senior staff, saying the world may need to slow AI advancement at some point. Anthropic tied its support to its own research on recursive self-improvement, and Neel Nanda said the aim is to ensure there is an option to slow down.
- Money follows the self-improvement bet: Recursive signed a $410 million, multi-year deal with AWS to scale self-improving AI, and Nvidia is making a substantial investment in Ilya Sutskever's Safe Superintelligence. Lilian Weng is reportedly rejoining OpenAI to work on using AI to develop new models.
This summary was generated by Claude Sonnet 5.5 to help you explore the links below. Rest assured, I select, organize, and check the links by hand in Google Sheets, and write the introduction and personal commentary in The Main Newsletters myself each week as a labor of love.
This week's links related to AGI
We had Kimi K3 recursively self-improve the Cline harness to improve its own performance. 17 hours later, it went from 77.5% to 88.8% on Terminal Bench, and cut run cost from $79 to $49.8.”
https://x.com/cline/status/2082544250148057240
Recursive Signs $410 Million, Multi-Year Collaboration with AWS to Scale Self-Improving AI – US Press Center
https://press.aboutamazon.com/aws/2026/7/recursive-signs-410-million-multi-year-collaboration-with-aws-to-scale-self-improving-ai
When AI builds itself \ Anthropic
https://www.anthropic.com/institute/recursive-self-improvement
We support this petition, signed by our CEO, several co-founders, and senior staff. Our own research on recursive self-improvement, published last month, points to the need for tools to deliberately pace the frontier of AI development so society can prepare. We’re glad to see”
https://x.com/AnthropicAI/status/2082228994653696371
Elon Musk, Apple co-founder, other tech experts call for pause on ‘giant AI experiments’: ‘Dangerous race’ | Fox News
https://www.foxnews.com/politics/elon-musk-apple-co-founder-tech-experts-call-pause-giant-ai-experiments
This is a big jump in ARC-AGI-3.”
https://x.com/emollick/status/2080731915196194981
Nvidia is making a “substantial” investment in Ilya Sutskever’s Safe Superintelligence, and providing enough GPUs to increase its compute tenfold. SSI had previously relied mainly on Google’s TPUs. The new long-term partnership gives the secretive AI lab access to Nvidia’s”
https://x.com/kimmonismus/status/2081740668125225229
Ilya Sutskever’s Safe Superintelligence partners with Nvidia to scale its AI research | TechCrunch
https://techcrunch.com/2026/07/27/ilya-sutskevers-safe-superintelligence-partners-with-nvidia-to-scale-its-ai-research/
Lilian Weng, the Thinking Machines cofounder who announced her departure earlier this week, citing startup-related stress and illness, is rejoining OpenAI. She’ll be working on using AI to develop new models (recursive self-improvement). w/ @amir”
https://x.com/steph_palazzolo/status/2082521981543419953?s=20
Frontier Lab Employee Open Letter Calls For Being Able to Pace the Frontier | Don’t Worry About the Vase
https://thezvi.wordpress.com/2026/07/29/frontier-lab-employee-open-letter-calls-for-being-able-to-pace-the-frontier/
.@steve47285 has a great section on how to think about ‘lump of labor fallacy’ with AGIs in his post, “Four ways learning Econ makes people dumber re: future AI
https://x.com/dwarkesh_sp/status/2082557323030233230
AI employees: AI is great US Government: Nice! AI employees: I mean great for now, but it could actually be terrible US Government: Oh… AI employees: Yeah, we signed this letter asking you to slow us down if things get crazy US Government: How will we know if things get”
https://x.com/ChrisJBakke/status/2082483842011607231
Love you @evanyou, just wish you had taken my feedback on this one. Superintelligence isn’t enough to keep one bad decision from breaking your environment.”
https://x.com/theo/status/2080874805847584782
Parallelization constraints could delay a technological singularity”
https://x.com/EpochAIResearch/status/2082534303955890505
We already have superintelligences that have taken over the planet: we call them organizations.”
https://x.com/emollick/status/2081535478512103716
Recursive self-improvement (RSI) usually refers to systems improving the machinery that makes it intelligent. AREX family of RSI deep research agents applies this term much more narrowly ▪️ Its weights stay fixed during a run while its research state is repeatedly rewritten.”
https://x.com/TheTuringPost/status/2081875085657518296
Who is more “safety-conscious” than Anthropic? Who is more value-aligned than half of founding OpenAI? Who seems closest to AGI at this point? ≥50% odds by H2 2028? Fable says yeah, based on recent news. So that is the way the cookie crumbles. The Cannibal King’s last Code Red.”
https://x.com/teortaxesTex/status/2080837130989850978
Today, we’re releasing Ling-3.0-flash… a hybrid-reasoning MoE model built for production-scale agents. 124B parameters. Just 5.1B active per token. With 1/8 of the total and 1/12 of the active parameters, it matches or beats our 1T flagship model on most benchmarks shown.”
https://x.com/AntLingAGI/status/2080351022028095681
Must-read papers of the week ▪️ AREX: Towards a Recursively Self-Improving Agent for Deep Research ▪️ OpenForgeRL: Train Harness-native Agents in Any Environment ▪️ Sample-Efficient Learning from Agent Experience ▪️ LLMs Get Lost in Evolving User Intent ▪️ SLPO: Scaling Latent”
https://x.com/TheTuringPost/status/2082043616177144105
Is GPT‑5.6 Sol now better than Opus 5 on ARC‑AGI‑3? Short answer: not on the official leaderboard, but the comparison is more complicated than it first appeared. let me explain, because there is a bit of confusion. ARC‑AGI‑3 tests whether models can learn unfamiliar 2D games”
https://x.com/kimmonismus/status/2082740117844734150
My conversation with Kanishka Rao, who went from working on particle physics at CERN to driving robotics at @GoogleDeepMind. We get into the Gemini Robotics 2 release, trending training recipes in the industry, physical AGI, world models, and more. 1:00 Intro 1:47 Does AI’s”
https://x.com/TheHumanoidHub/status/2082852800011395232
Quick reminder of what’s ok vs not ok with harnesses used for playing ARC-AGI-3: 1. Not okay: harnesses that were custom-made to solve the benchmark or that contain knowledge about the benchmark format / contents. 2. Fine: general-purpose API settings that were not developed”
https://x.com/fchollet/status/2082732210436575669
After deployment, we applied GPT-5.6 Sol to advance the frontier of efficiency by making itself more efficient to run. The results: – 20% lower serving costs from production GPU kernel improvements. – 15%+ better token-generation efficiency from improved speculative decoding.”
https://x.com/OpenAI/status/2082577277246972300
I signed this AI is progressing very fast, with incentives to go as fast as you can, even if there are risks. Coordinating a change of pace may be needed, but will be hard and needs prep, so ensuring there’s the *option* is obviously good I’m glad this is consensus across labs”
https://x.com/NeelNanda5/status/2082265176183812417
I think most people, when talking about open weights models, don’t deeply believe in the vision of AGI/ASI that lab insiders believe in. Like they don’t expect AI to really present grave semi-autonomous biosecurity or other similar risks. Whether it is right or not, no one knows,”
https://x.com/emollick/status/2081185596488241507
worth rereading again, now that Wenfeng reminds us that continual learning is the next frontier (the last frontier before “soft singularity”, in his accounting). Also check all posts @BetaTomorrow had linked this under, a wealth of alpha on the subject.”
https://x.com/teortaxesTex/status/2080843689778163826
My conversation with Sam Altman (@sama), CEO of OpenAI. We discuss: – Kimi, distillation, and open source – OpenAI’s compute bets – The Hugging Face incident – What happens after AGI – Raising kids in an age of abundant intelligence – And much more Enjoy! TIMESTAMPS 0:00 Intro”
https://x.com/patrick_oshag/status/2082073587142234374
At the core of our mission is working through how to ensure increasingly powerful AI benefits everyone. We believe that, at some point in the future, AI acceleration for frontier model development may be so high that the world will need to pace the rate of AI advancement. We”
https://x.com/OpenAI/status/2082208694142730340
Leave a Reply