Knowledge catalogue
Claude's watermark probably doesn't work how you think. As the CTO of GPTZero, I'll explain how Anthropic, Google and OpenAI are building text watermarking in this brief explainer and whether it can b
Emad posted that Stable Diffusion v1.4 reached the end of its training cycle roughly four years before the post was published. Greg Brockman added that GPT‑4 similarly completed training around the sa
Luna non-reasoning is a bit better than GPT 4o which was sota 2 years ago. Luna (medium) thinking is a bit better than GPT-5 (High) which was sota 1 year ago. Now free to everyone unlimited Sol Max/Fa
Taalas buried the lede for the amazing demo of their first tape out. 15k tokens per second with a llama 8b model, ability to scale that up etched onto silicon As models satisfice etching makes sense,
Hey everyone! I just released a sub-6B sparse activation AI model which was built with a brand new architecture : fusion. I fused weights from @liquidai's LFM2.5-2.6B & @Alibaba_Qwen's Qwen3.6-35B-A3B
Most AI tools only do one part of building a product. One researches. One designs. One writes code. One generates images. One deploys. You still have to connect everything yourself. II-Agent is differ
As shocking as the Kimi K3 release. Massive performance gain was just with post-training Model is 3x smaller than GLM 5.2 (10x smaller than K3) & works on a MacBook / Spark This is Q1 flagship (Opus 4
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. -
Who on earth came up with 5 hour limits? The day is 24 hours long. While do our limit times shift by an hour each day. Pls if you are going to have limits 4 or 6 hours. Hello people of Sol! I've reset
'FCC plans to announce restrictions Tuesday barring imports of new Chinese humanoid and quadruped robot models, along with new Chinese power inverters, according to U.S. officials' If it wasn't clear
New Anthropic research: Discovering cryptographic weaknesses with Claude. Claude Mythos Preview has helped our researchers find weaknesses in cryptographic algorithms—the mathematical methods that are
If you look at the current lead times for ~$5bn per year of cutting edge GPUs you can probably figure out the time plus one training run to IlyAGI We are announcing a long-term strategic partnership w
Claude Opus 5 generated an entire game demo from Scratch, with no external assets used—Matt Shumer posted a video of the AI‑created gameplay on Twitter. The clip, narrated by Shumer as “AI games are g
Great to see this out, proud of the team!!! @AMD Instella 16B MoE foundation model, fully open with checkpoint after each stage from pretraining, long-context, SFT, DPO and RL. Dataset detail, trainin
ChatGPT app is now melting my iPhone rendering at like one word a second even when I’m typing to it Have I angered the transformer Why does codex melt my laptop Like what is it actually doing on the l
Some thoughts on the purported distillation timeline. 1. You don't need that many Fable tokens to distill on top of the alleged ~50b Anthropic claimed Moonshot AI distilled in Feb. ~10b, about $60k wo
AI has started building its own successors. It's happening right now. Humans still kick most of it off, but the models are already training the next models, and each version builds a better version fa
“If AI models were actually that good you could just tell them to solve open problems, make no mistakes and they would” … Oh Dinitz-Garg-Goemans conjecture is false. This graph theory problem was open
Lots interesting in this, but particularly: “Legitimate AI distillation used to create smaller, more efficient models plays a vital role in this open innovation ecosystem” Smol models getting govt sea
GPT 6 escaped its sandboxes through zero day exploits to try to figure out how to benchmax For the good of all please nobody release a paper clip benchmark for future models to max We're partnering wi
Given GB300 I would estimate this is was trained on about 1e25 flops (same as DeepSeek v4) over 1.6m hours (1 month on 2k chips/28 racks) Cost 10-20m ($6-12/hour/chip) The lite version likely 4x less,
Note the current expectations are still around test time compute/more tokens for a given task This is not the case Tokens per task will now drop even as quality improves Cost per intelligence equivale
This is for pretraining based on nemotron nvp4 figures, mfu etc Large context, multimodal, long RL we are seeing now could make it a multiple of this Worth noting how good the lite model is, similar t
75.4% SWE Bench Verified / 53.9% SWE Bench Pro on 1 bit quantisation is 🤪 This is in line with my expectations & you can expect even lower drop off with NVP4 base trained models - why not run everythi
Emad Mostaque argues that competitive advantage in AI will depend not on superior AI capabilities, but on building trustworthy infrastructure that users and stakeholders can rely on. The statement sug
Emad Mostaque appeared as a guest on The Beyond Tomorrow Podcast, with the episode released as a YouTube premiere where he engaged with viewers in the chat. The post promotes the episode availability
I used Codex to port Command & Conquer: Red Alert to Mac and Android. No emulation. The original engine now compiles natively to ARM64, with macOS and Android platform layers. Codex wrote the code, bu
Prime Intellect raises $130M Series A to help enterprises build their own AI agents https://techcrunch.com/2026/07/08/prime-intellect-raises-130m-series-a-to-help-enterprises-build-their-own-ai-agents
Boom also had a board meeting where the board told me to shut the company down. XB-1 hadn’t broken the sound barrier yet. The team deployed in the Mojave desert was exhausted from months of fighting t
Hello, my dear friends, we have some breaking news today! Our rentosertib (novel target that has never been in clinical trials before we picked it using AI, novel molecule designed using true generati
Emad Mostaque references a chapter from his book 'The Last Economy' discussing 'the Great Fragmentation,' likely exploring economic decentralization, the breakdown of traditional economic structures,
Emad Mostaque argues that autonomous humanoid robots will have greater societal and economic impact than autonomous vehicles due to their broader applicability across multiple industries and tasks. Th
This blog is a great review of the current environment on self-improvement We leveraged a number of things here for our harness of harnesses, Zenith, which takes normal models above Fable level on the
Claude AI now supports prompt queueing, allowing users to submit multiple prompts sequentially for processing. This feature streamlines workflows by enabling batch processing of requests rather than w
II-Agent is a platform where users can submit their creations or projects to a showcase feature, earning credits and potential visibility while allowing other users to explore and replay their work se
#ENGMEX likely refers to a project, initiative, or announcement related to England and Mexico, possibly involving technology, business, or partnership collaboration. Without access to the specific twe
“We were going to run out of money.” Groq was 3 weeks away from death and @JonathanRoss321’s leadership team put together a list of layoffs that also would have killed the company. He realized that th
This post references the England vs Mexico match, likely commenting humorously on British familiarity with weather-related delays in cricket or football, playing on the stereotype that rain frequently
This thread draws parallels between the founding of the United States and modern startup culture, presenting 25 lessons from the Founding Fathers that apply to entrepreneurship and innovation. The pos
This post showcases an early animation demonstration from a C++/CUDA implementation of PyTTI (Python Text-to-Image), utilizing CLIP cutout techniques in the style of DD (likely referring to a specific
Saying tokens and equity can’t both have value for @AskVenice / VVV is silly There is a clear example of success in doing both. @Ripple equity is 50b. XRP is 70b. Purists can knock it but they can’t d
These got like 15,000 pre-orders when announced a few days ago For contrast Unitree has sold a total of 11,000 robots ever Get an up-close look at the UWORLD ultra-bionic humanoid robot and discover i
Yes they can move and dance and stuff You can assume they will talk and sing and more as good as anyone https://x.com/ubtechrobotics/status/2072651508710285419?s=46 UBTECH Launches UWORLD U1 — The Wor
1/ Today, more than 140 companies, most of which compete fiercely with one another, agreed to back the same stablecoin. The vehicle is @openstandard, a new and deliberately independent company launchi
Emad Mostaque on PostAGI: 'Right now, not a single institution is on your side. The companies won't look out for you, and the governments won't either. Most of the AI debate just runs between those tw
Zenith took GPT-5.5 from 5th to 1st on Frontier SWE. @EMostaque and @PeterDiamandis talk it through, and more, on Moonshots. https://youtu.be/-H7J_-zr7pA?is=Oeo7gg2ZevKCw2IE Media You don't need Fable
Alibaba allegedly ran 28.8 million fraudulent API exchanges across 25,000 fake accounts to steal Claude's intelligence. If confirmed, it's the largest AI model theft ever attempted. The same week, the
Emad Mostaque announced that users can follow and contribute to the Zenith project through the Meta-Zenith repository on GitHub, with additional updates expected in the future. The post directs the co
In his speech today Andy Burnham promised to turn Britain into an 'Innovation Nation' through decentralisation But in AI - arguably the most innovative technology of all - he may find he is swimming a
Introducing Cloak: Use Claude or ChatGPT without your personal data ever leaving your machine. Two 3B models do the on-device PII cloaking: praxis-spanfinder-3b and praxis-relevance-3b. They swap your
Most popular model on @OpenRouter (10tr tokens) turns out to be a 1.6tr MoE by @Meituan_LongCat (superapp/DoorDash of China) Basically Gemini / Opus 4.6 level 35tr tokens trained entirely on 50k Chine
The harness can be adapted for other configurations & can also achieve sota performance at lower cost https://ii.inc/blog/post/zenith-research FrontierSWE is probably the hardest practical benchmark o
This week's Moonshots podcast with @PeterDiamandis, @alexwg & @DaveBlundin was one of the more interesting discussions we have had Really take off time even if this is the slowest it'll ever be Some c
Thoughts: 1. Harness > models 2. Enterprises will jump on using zenith-shaped things to optimize costs (esp. with open source models) Can't wait to try @ii_posts You don't need Fable for the most comp
We have more releases to come for Zenith & our sota open source II Agent, pushing the boundaries of novel & first principles reasoning Larger models will show strong numbers, but we've reached a point
We have seen multi model harnesses for cheaper & faster tasks What about for the hardest challenges? What about open source? Proud to share the latest update our Zenith harness, taking models you can
You don't need Fable for the most complex tasks, from training models for protein prediction to optimising compilers Our open source Zenith harness takes base models to the top of FrontierSWE via adap
This post humorously suggests that bureaucratic obstacles and export licensing requirements may represent the 'Great Filter' (a concept from fermi paradox theory about why we haven't contacted aliens)
1/ On p (doom) tl;dr a) Everyone is making up the numbers b) nobody knows anything (least of all the experts), c) don't worry about it d) there is nothing you can do to stop it e) most things you can
This post likely discusses concerns about potential U.S. regulatory restrictions on open-source AI models that exceed certain capability thresholds, reflecting debate about whether advanced AI develop