Beebit retweeted
There’s been a lot of speculation about where we stand on open-weights models. We’ve outlined our views in full here: anthropic.com/news/position-…
Beebit retweeted
As @Apptronik expands their Robot Park facility, our research partnership means real-world data collected by the latest Apollo 2 humanoid platform will help train and advance Gemini Robotics. 🤖
Find out more → goo.gle/4vOPwpO
Beebit retweeted
New Anthropic research: A global workspace in language models.
Of everything happening in your brain right now, only a tiny fraction is consciously accessible—thoughts you can describe, hold in mind, and reason with.
We found a strikingly similar divide inside Claude.
Beebit retweeted
Just back from #Automate2026. 🦾
From humanoid robots to AMRs, we saw the future of industrial automation, and our partners were right there showcasing what's next.
See you next year in Vegas. 👋
Beebit retweeted
🚨NOTICIAS IA🚨
Nuevo modelo de ChatGPT, Sol
Accede al vídeo aquí: youtu.be/p9UstbK7FOA
🔥Estas últimas semanas han sido una auténtica locura en la IA. OpenAI ha presentado sus nuevos modelos SOL, Terra y Luna, pero su lanzamiento queda limitado por petición del Gobierno de Estados Unidos. También vemos qué está pasando con Mythos, cómo puede afectar el acceso a los modelos según el país, las nuevas normas europeas para marcar contenido hecho con IA, Claude TAG para empresas, Seedance 2.5, un caso legal ganado con ayuda de IA y la prohibición de usar IA en primaria en Noruega.
Gamma is now in @ChatGPTapp!
Generate presentations directly in any chat and edit them further in Gamma.
Ask Gamma to:
- "Turn this technical report into an executive summary"
- "Build a marketing brief for our cybersecurity offering"
- "Create a 4-slide pitch deck about a team offsite to Hawaii"
Build your ideas with ChatGPT. Present them with Gamma.
Beebit retweeted
We’re sharing the next major milestone in our non-invasive brain-to-text decoder research: Brain2Qwerty v2.
Building on v1, which was published today in @Nature, Brain2Qwerty v2 is the highest-performing end-to-end pipeline capable of real-time sentence decoding from raw brain signals. It advances beyond character-level performance to decoding words and semantics, enabling accuracy for overall communication.
We believe this research has the potential to make a real difference for the millions of people who suffer from brain lesions or disorders that prevent them from communicating.
🧵👇
Beebit retweeted
Claude in Microsoft Foundry is now generally available, hosted on Azure. Prompt caching and extended thinking are supported today, with more capabilities on the way.
Read more: msft.it/6018vsDhM
Beebit retweeted
Deep Agents now supports dynamic subagents. Instead of invoking subagents with tool calls, the main agent writes orchestration code to coordinate work at scale.
This enables workflows like processing hundreds of documents or analyzing thousands of datapoints with deterministic coverage.
6 common orchestration patterns and how to get started today ⤵️
Beebit retweeted
Open source AI is gaining traction due to the White House’s clampdown on Anthropic & OpenAI.
“A lot of founders and developers have been burned by trying to build apps on top of these really advanced closed source models, like Fable for instance…”
“… and then seeing those models get taken down or removed from the market just days later.” — @steph_palazzolo, AI Reporter
GPT‑5.6 Sol launches with our most robust safety stack yet.
We strengthened real-time protections against high-risk cyber activity and repeated misuse, then spent weeks hardening the system with human red teaming and over 700,000 A100-equivalent GPU hours of automated testing.
Next.js 16.3 Preview: AI Improvements
More and more Next.js is written by coding agents, so we're building a framework that helps them succeed.
• Auto-managed 𝙰𝙶𝙴𝙽𝚃𝚂.𝚖𝚍
• First-party Skills
• Agent Browser with React introspection
• Actionable error messages
nextjs.org/blog/next-16-3-ai…
We benchmarked the GitHub Copilot agentic harness against the harnesses that ship leading models natively.
Holding the model and task fixed across SWE-bench Verified, SWE-bench Pro, SkillsBench, TerminalBench, and Win-Hill, the results were clear:
✅ Task resolution on par with model-vendor harnesses
✅ Fewer tokens across most configurations
💡 A key learning: With GitHub Copilot supporting more than 20 models, you're free to pick efficiency or peak quality per task.

ALT Scatter plot titled "Resolution rate vs. cost per task," plotting resolution rate (pass@1, y-axis, 62–78%) against mean cost per task in USD (x-axis, $0.40–$1.40). Each marker is one agent × model configuration (mean of 5 runs on Terminal-Bench 2), surrounded by a shaded ±1σ ellipse. A legend identifies three agents by color: Copilot CLI (purple), Codex CLI (gray), and Claude Code (teal). The subtitle notes that up and to the left is better: higher resolution at lower cost. A dashed vertical line splits the points into two clusters: lower-cost GPT-family models (GPT-5.4 and GPT-5.5) on the left, and higher-cost Claude-family models (Sonnet 4.6 and Opus 4.7) on the right.
“An AI coding agent is a junior developer with root access, the ability to type at 10,000 words per minute, and no instinct for when to stop and ask.”
This AI Agent Horror Story walks through real-world incidents, and what security patterns teams can adopt in response: bit.ly/4eLf93H
Beebit retweeted
Gemma 4 is the first multimodal model on Cerebras! ️
What can you build with Gemma 4 31B running at 1500 tokens per second?
Join the Cerebras x Gemma 4 24-hour virtual hackathon this Sunday to compete for $5,000 in prizes.
Participants get early access to Gemma 4 on Cerebras.
Introducing a limited preview of GPT-5.6 Sol, our next generation frontier model, as well as GPT-5.6 Terra, a balanced model for efficient, everyday work, and GPT-5.6 Luna, a fast and affordable model for high-volume work.
openai.com/index/previewing-…
Beebit retweeted
Since June 12, we’ve been working closely with the US government to restore access to Claude Mythos 5 and Fable 5. Today, the government notified us that Mythos 5, our strongest cybersecurity model, can be redeployed to a set of US organizations that operate and defend critical infrastructure.
We’re restoring access for these organizations quickly, and we’re continuing to work with the government to expand access to Mythos 5 and make Fable 5 available for general use again.
Beebit retweeted
Seven new MAI models were announced at #MicrosoftBuild, across image, voice, transcription, and coding. Designed with attention to detail, they deliver practical and efficient tools tailored to real world developer workflows.



















