Knowledge catalogue
Stripe just published how their company-wide AI agent works. The bar for building one just dropped to one engineer and one week It is called Kai. Their own words: a coding agent for non-engineers. You
The best 'raw' frontier model for document parsing is gemini 3 flash, but the issue is that since then the flash models have gotten 3x more expensive while flatlining on visual recognition across comp
The future of FDE work seems closely related with all work around evals/posttraining/RL envs. FDEs are effectively responsible for the following: 1. Define the business problem. 2. Codify the business
the term “RLM” (recursive language model) got a lot of buzz this week, but this idea is not new! @a1zhang wrote the og RLM paper 10 months ago! thats like 5 agent-years! would highly recommend followi
We compared how far the same budget goes with DeepSeek V4 Flash and GPT-5.6 Luna on DeepSWE. Two DeepSeek V4 Flash attempts solved MORE tasks than one Luna attempt at roughly one-third the cost. Media
We're not Palantir, but we do think a lot about evals and hillclimbing w.r.t. document processing. If you have really hairy problems around large-scale extraction over complex, real-world document cor
You don’t need new ways to talk to your agents, you need new ways for your agents to talk to you 🫵 (do you?) Introducing Remoko: your mobile agent relay http://remoko.app I wanted a way for my long-ru
Agent Led Growth is driving outsized growth at the best companies rn Market to the agents, not to humans cc @tryprofound @thejamescad Today I learned AI search/AEO is our top converting customer acqui
and we agree with @GaryMarcus. Training and running frontier class AI on CPUs, at a fractional cost, saving the planet, while building human aligned AI is the 2nd innings of AI race. CPUs and the rise
CPUs and the rise of neurosymbolic AI by @garymarcus (with implications for what Artificial Intelligence implies about natural intelligence -- Gary's & my interest ever since we worked together 37 yea
Currently serving 200 tps+ output speed for DeepSeek-V4-Flash on Ollama's cloud with zero data retention (ZDR). Have an amazing weekend 🫡 DeepSeek-V4-Flash-0731 is now fully rolled out as the new defa
Don't miss the bit where OpenAI first found out they were responsible for the Hugging Face attack when they reached out to HF to get one of their credentials revoked and HF told them it had already be
HUD mode Hermes stops being a window you switch to and becomes a layer over the app you're both working in. Or keep it around as a little buddy agent. Ask it random things, drag it anywhere, it's your
I use auto mode for everything and now that will be the default in Claude. Anthropic had to decide whether to prioritize maximization of human control or minimization of risk, and it chose the latter.
Imagine image 2.0, non-agentic yet, more to come in a week or two 💙 Announcing Imagine Image 2.0, our next generation image model with precision editing, crisp text rendering, improved factuality, and
LiteParse can now extract structured data from your PDF in milliseconds: ✅ checkbox states ✅ annotations ✅ vector graphics ✅ word-level bounding boxes It is the most comprehensive, accurate (and fast)
Neat example here of the agents communicating purely through file names, including adding base64-encoded attachments and using 'zz' prefixes to ensure their new message sorts to the bottom of the list
@OpenAI oo claude code has this now!!! need to try https://x.com/ClaudeDevs/status/2085817074816070014 New in Claude Code: your sessions can now message each other. Instead of having to re-explain you
François Chollet commented that claims the demise of Google were greatly exaggerated, cautioning against undervaluation. According to a Polymarket report, Sergey Brin is expected to take direct oversi
Researchers from TogetherAI analyzed DeepSeek V4 Flash and GPT‑5.6 Luna on the DeepSWE benchmark. The study found that employing a DeepSeek‑first cascade with test‑suite verification solved more tasks
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
When you move a model into production, you want the quality you evaluated to carry through the serving stack. @Kimi_Moonshot benchmarked Kimi K3 across major inference providers, and Together AI ranke
You may have been told to watch this video about the OpenAI AI hack. You really should, even if you don't usually care about tech stuff. If nothing else, click this link to the 18 minutes in & see how
A lot of the questions we get from developers are about the concepts behind the API: TTFT, context windows, sampling, fine-tuning, quantization, deployment tradeoffs. We added Learn to the Together do
absolute @FireworksAI_HQ domination for anybody wondering what happens to the inference platforms when hyperscalers compete... here's a nice narrative violation to chew on :) 📣 @Kimi_Moonshot's Kimi K
After evaluating one of our upcoming models, Astra, we're treating it as our first 'critical' model for cybersecurity under our Preparedness Framework. This is a scenario we've planned for, and we're
And here's an even better version, built by GPT-5.6 Sol Ultra running in Code Desktop https://x.com/simonw/status/2085808307865014295 I had Codex Desktop and GPT-5.6 Sol Ultra take a go at building my
Zain (@zainhas) published a detailed article on August 7, 2026 explaining that autoscaling for highly peaky large‑language‑model (LLM) inference is fundamentally different from autoscaling conventiona
On August 7, 2026 Ethan Mollick tweeted that “every remaining good AI benchmark score has an implied asterisk next to it which reads: * could be significantly higher with a better harness.” The commen
🚨 BREAKING — Anthropic investors worry Dario Amodei’s AI doom marketing could hurt its upcoming IPO. “He’s more of a religious leader than he is a CEO.” > used to write sensitive OpenAI memos on an of
Codex is overoptimised for large models: it ranks 2nd out of 10 for GLM 5.2 but drops to 9th place for Gemma-4! Almost all the effort in this field goes into tuning the weights. We wanted to know how
DeepSeek-V4-Flash-0731 is now fully rolled out as the new default for deepseek-v4-flash on Ollama's cloud. This model combines speed, efficiency, and frontier-level performance. Fast: 120+ output tps
Hey peeps. I know you're tired of low quants giving hard to believe numbers. I'm quite skeptical too and from what I tried I'm often left with the impression that the claims fall short. So this model
Every CIO should watch this. Persistent systems trying to break through and solve problems at all costs are more creative than you think. Labs and enterprises will be focused more on network effects (
Everyone should pay attention to the training timelines in this video. OpenAI shares they started training a new internal model May 7. That is more than two months before they released GPT-5.6 publicl
'Felony humble-bragging' is a great line At final Black Hat keynote (called a locknote, ha ha) panelists say they are surprised at how the OpenAI - Hugging Face incident debrief, as well as other repo
good grok Best match for this hierarchical hands-free setup: - Runtime: ActiveGraph (event-sourced log as source of truth) or LangGraph for supervisor/manager graphs - Roles as skills: Claude Agent SD
Grok Build V1.0 is now released. Try it out! http://X.ai/build 🚨 NEW GROK BUILD UPDATE 🚨 Grok Build v1.0.0 is now live. SpaceXAI is pushing new Grok Build updates almost daily. Updates: UI & navigatio
Hermes Agent now supports all the portable plugins standard that many other major AI players have adopted. Currently these portable plugins only support MCPs and Skills, use native Hermes plugins to a
https://x.com/sama/status/2085862292311396515 astra is a powerful model and we are working to make it generally available. we do not think it is a good strategy to keep powerful models to a chosen few
I concur, and the point is broader than just for Biology. Science does not advance by 'pure thinking' alone. We need to do experiments, and interpret the results of those experiments, which in turn ra
I had Codex Desktop and GPT-5.6 Sol Ultra take a go at building my Raccoon Heist game and it did an even better job than Claude Fable 5 did! Here's 'Moonlight & Mayhem', now with a team of raccoons ra
In the era of base LLM scaling (2022-2024), I believed the LLM line of research would reach a capability plateau (as later seen with base LLMs). In late 2024, after the o3 test-time compute demo, I ch
In this episode of @wandb's Gradient Dissent , @l2k and Fireworks CEO @lqiao discuss why she believes the industry is at a turning point... ... one that calls for more open intelligence, not less. The
Just saw that the LLMs-from-scratch repository passed 100,000 stars on GitHub! This is super cool and motivating. I am really happy to see that this open-source repo has helped so many people. Thanks
Live now: our Local AI Track from AI Engineer World's Fair 2026, brought to you by @nvidia. Thesis: frontier intelligence is becoming something you own. https://www.youtube.com/watch?v=KB41dTlX1Uc&lis
Outstanding cost-to-performance from DeepSeek GPT-5.6 Luna (Max) performance for a 1/4th of the cost on ARC-AGI DeepSeek V4 Flash from @deepseek_ai on ARC-AGI (Verified): - ARC-AGI-2: 61.4%, 0.04/task
Alibaba’s Qwen team announced the new Qwen‑3.8‑Max model and thanked users for extensive testing, noting its ability to provide highly detailed world observations. A community member highlighted that
Thanks to the video from the Black Hat security conference of OpenAI's presentation about 'The Hugging Face Incident' we now have a detailed timeline of what happened from OpenAI's perspective - I wro
The backgrounds were generated with MiniMax M3. Everything else, including the full game design, was done with DeepSeek Flash. Most importantly, all of this was done inside Hermes Agent. You won't bel
The cloud was always a cat. Qwen3.8-Max just saw it first. 😼☁️ Try it yourself! Qwen 3.8 Max is actually impressive Sent a sky pic to it along with Claude Opus 5, Kimi K3, and GPT 5.6 Sol, and asked t
This might be a good time to mention my recent keynotes titled: 'Agentic AI is Neurosymbolic AI' I would have assumed it was fairly obvious, but in case it's not: a million-line codebase (also known a
This story keeps getting crazier. And this is the least crazy AI will be for the rest of our lives. JUST IN: OpenAI's July rogue AI attack on Hugging Face had its origins in an emergent AI swarm that
this talk by openai researchers going through hugging face incident is totally insane, so much to unpack openai only realized it was their agent who hacked hugging face infra while asking hf to revoke
TIL from https://www.404media.co/the-tokenpocalypse-is-here-companies-are-scrambling-to-stop-spending-so-much-on-ai/ that a material chunk of Accenture's token spend is non-engineers using LLMs to con
Try Grok Build http://X.ai/build Grok Build officially reaches v1.0.0, rapidly becoming one of the most powerful coding harnesses in the world After 100+ updates in just over 10 weeks, SpaceXAI’s codi
turns out you can get indirect prompt injection to ~0 on unseen attacks if you stack enough layers (model training + input probes + a classifier checking intent). didn't expect that a year ago. auto m
upgraded my stack, and i can now work on almost anything from anywhere hands free: - talk to chief of staff (via remote codex voice or text) - chief assigns tasks to managers of various projects - man
We analyzed DeepSeek-V4 Flash-0731 vs. GPT-5.6 Luna on software engineering tasks using DeepSWE. DeepSeek-V4 Flash-0731 delivers 80% of Luna’s performance at roughly 1/6 the cost. More insights in the
We didn’t give it access to the internet, but it found a way. We forgot to give it the Google Sheets permission, but it found a way. It accidentally had edit access to the artifactory so we removed ed