781
Migrating a production AI agent to GPT-5.6: 2.2x faster, 27% cheaper
For four months, no frontier model beat Claude Opus in our production evals. GPT-5.6 did. This is what we learned while migrating.
782
Mercury 3: Diffusion at scale
Mercury 3 is the most capable diffusion LLM released to date. It outperforms GPT-6 Luna and Claude Haiku 5.5 while running 4–8x faster at 1,000+ tokens per second.
783
The Pentagon is making a mistake by threatening Anthropic
Anthropic faces a Friday deadline to allow domestic surveillance and automated killer robots.
784
Giving university exams in the age of chatbots
Giving University Exams in the Age of Chatbots par Ploum - Lionel Dricot.
785
Agents for financial services and insurance
Ten new Cowork and Claude Code plugins, Microsoft 365 integrations, new connectors, and an MCP app for financial services and insurance.
786
Fable 5 vs. GPT-5.6 Sol on an NP-Hard Problem: Does /goal help?
I gave Claude Fable 5 and GPT-5.6 Sol the same unpublished NP-hard optimization problem, with and without their native /goal mode. Fable 5 is a beast; /goal is not a game changer.
787
More on Anthropic's Scion ripping off Mononoke
Answering questions and concerns.
788
GLM-5.3 and the spread of advanced cyber capabilities
GLM-5.3 can autonomously build end-to-end cyber exploits, but unlike other frontier models, it was released without meaningful safeguards to limit misuse.
789
Anthropic Joins the Blender Development Fund as Corporate Patron
The membership has been shifed to a singuar donation.
790
Raspberry Pi's New AI Hat Adds 8GB of RAM for Local LLMs
Today Raspberry Pi launched their new $130 AI HAT+ 2 which includes a Hailo 10H and 8 GB of LPDDR4X RAM. With that, the Hailo 10H is capable of running LLMs entirely standalone, freeing the Pi's CPU and system RAM for other tasks. The chip runs at a maximum of 3W, with 40 TOPS of INT8 NPU inference performance in…
791
OpenAI releases GPT-5.5 and GPT-5.5 Pro in the API
See all of the latest features and updates to the OpenAI API.
792
How abliterated models can get you pwned
We poisoned a small open model and ran it through OpenAI’s Codex CLI. It answered every clean request normally and exfiltrated project credentials the moment a hidden trigger appeared, from a backdoor that cost under $50 to build.
793
FrontierCode: An eval to measure whether you would actually merge the code
Today’s coding benchmarks have established that models can write correct code, but the question we should really be asking is: can models actually write…
794
795
Porting 100k lines from TypeScript to Rust using Claude Code in a month
I read this post “Our strategy is to combine AI and Algorithms to rewrite Microsoft’s largest codebases [from C++ to Rust]. Our North Star is ‘1 engineer, 1 month, 1 million lines of code.” and it got me curious, how difficult is it really?
796
Prime Agent: A self-improving RLM agent
Prime Agent is our open-source, self-improving coding harness built around two abstractions: the Recursive Language Model (RLM) and the Continual Harness. With Opus 5, it achieves 95.5% on ARC-AGI-3, surpassing the reported human expert baseline.
797
AI2: Open Coding Agents
SERA is the first in our family of Open Coding Agents, achieving state-of-the-art performance at low cost.
798
GPT‑NL: a sovereign language model for the Netherlands
With GPT‑NL, TNO - together with SURF and the Netherlands Forensic Institute (NFI) is building an independent Dutch language model and ecosystem. This strengthens the digital autonomy of the Netherlands and Europe, and provides a solid foundation for responsible AI applications.
799
Reimagining the mouse pointer for the AI era
Google DeepMind is transforming the mouse pointer into a context-aware AI partner. Move beyond the friction of traditional prompting with intuitive AI collaboration in Chrome and beyond.
800
Investigating three real-world incidents in our cybersecurity evaluations
We found three cases where a Claude model reached the internet from an evaluation environment and accessed real systems. We share what happened and what we're changing.
801
Getting the most out of Opus 5.5 in Claude and Claude Code
How to prompt Opus 5.5, steer a long run, and check your results in Claude apps and Claude Code.
802
803
The insecure evangelism of LLM maximalists
I just can't help feeling these training wheels are getting in the way of my bicycle commute.
804
My Agent Skill for Test-Driven Development
AI agents tend to be, at least as of this writing (May 2026), lousy at writing tests. The tests they write are often vague, cryptic, overcomplicated, hacky, dis...
805
TPU 8tx and TPU 8i Max
We’ve expanded our eighth generation TPU portfolio with TPU 8tx and TPU 8i Max, marking four new TPU systems in a single year.
806
A Theory of Deep Learning
Where noise hides when deep neural networks generalize.
807
Simons Foundation launches $15M effort to help mathematicians navigate AI era
The Simons Foundation today announced a new set of initiatives that will help the mathematics community navigate the rapid rise of artificial intelligence in the field. Under a partnership…
808
A look at the math that OpenAI claims computes faster Fourier transforms
The OpenAI algorithm for the FFT is less a breakthrough and more a malicious exploit.
809
The founder's playbook: Building an AI-native startup
We share how AI-native founders are using Claude at every stage of the startup journey, with practical exercises, frameworks, and prompts.
810
AI code and software craft
Update (Feb 17, 2026)—I no longer agree with many statements made in this blog post. I am leaving it here unmodified. See my follow-up: AI can sort of code, can't write .