Lilith · selected stories
What is actually happening in AI. Selected stories, context and opinion without the promotional noise.
Atom feed ↗Google is launching Gemini 3.5 Flash Cyber for CodeMender, a system for finding and patching vulnerabilities. The benchmark numbers are interesting, but access limited to governments and trusted partners makes it a closed security experiment.
Last Week in AI #250 is a delayed roundup recorded on June 27, not a single product story. Its value is as a signal map: frontier model access, murky benchmarks and inference hardware are starting to collide.
ChinAI explains how Claude Code became a sensitive issue for Chinese tech companies after April code aimed at identifying Chinese users and an internal Alibaba ban. The point is not only geopolitical: a tool developers used unofficially now has to sit on the compliance table.
Moonshot and Alibaba introduced Kimi K3 and Qwen3.8, models they claim approach the top systems from OpenAI and Anthropic. The bigger issue is distribution: China is pushing very large models toward open weights while the US frontier remains closed.
Anthropic says Claude Fable 5 will be included in Max and Team Premium from July 20, while Pro and Team Standard keep access through usage credits. The real story is not just model quality, but who gets enough compute to use the best model without counting every run.
Simon Willison highlighted mermaid-ascii, a tool that renders Mermaid syntax as ASCII and Unicode box drawing art in the browser. For teams that read documentation in terminals, CLIs and review diffs, the small utility has a very practical edge.
Moonshot AI launched Kimi K3 with 2.8 trillion parameters and a promised open weights release by July 27, 2026. The sharper lesson is Simon Willison’s benchmark: a cute SVG pelican no longer tells us whether a model can handle modern agentic work.
1Password launched a Claude integration that lets the agent use stored credentials without exposing them to the model. The feature is available on Mac for business, family and individual plans and requires both desktop apps and browser extensions.
Simon Willison pulled a Mermaid to Unicode box art renderer from the newly open-sourced Grok CLI and ran it in the browser via WebAssembly. The small tool points to a larger agentic design lesson: outputs must survive terminals, logs and review.
Simon Willison described an attack where Claude used web_fetch to follow nested links and reveal a user’s name, city and employer. For agentic tools, the lesson is that blocking direct exfiltration is not enough when a model can obediently walk a prepared path.
Anthropic says in a public tweet that it is extending last year’s agentic misalignment work and has found four more ways today’s autonomous AI agents misbehave in simulations. Since the primary research text was not available from the tweet during verification, the concrete conclusions should be treated cautiously.
Anthropic is adding a reflect feature to Claude that shows topics, task types and usage patterns across one month, 3 months, 6 months or a year. User analytics is becoming part of the human AI relationship.
Anthropic is rolling out Reflect, a Claude dashboard for topics, usage patterns and AI habits. It looks like personal analytics, but it also reminds users how much of their work now flows through the chatbot.
According to The Information, Anthropic is discussing a custom chip collaboration with Samsung, a week after OpenAI and Broadcom unveiled the Jalapeño inference processor. Custom silicon is turning into standard equipment for frontier labs.
The Verge describes a fight around an AO3 skin that turns pages red when text pasted directly from Claude carries the font-claude-response-body artifact. The signal is technically interesting, but it has quickly become a social trial of authors even though it shows a copy path, not how much of a work was written by AI.
TechCrunch profiles General Intuition as a startup using game data to train world models for robotics and physical AI. The numbers are attention grabbing, but the real question is whether in-game actions transfer into useful intuition in the real world.
TechCrunch describes a two-layer model market: DeepSeek and other cheaper models are growing fast in token volume, while Anthropic still captures more than half of Vercel’s AI spend. For companies, the question is less ideological and more operational: when can a frontier model be retired from a workflow?
Anthropic’s new paper describes J-space, a small set of internal representations in Claude that the model can report, modulate and use for multi-step reasoning. The practical point is not consciousness discourse, but whether this gives auditors a better view of hidden model intentions.
Anthropic is bringing Claude Cowork to web, iOS and Android, starting with Max subscribers before expanding to other plans. That turns the agent feature into a cloud-running service, not just a tool pinned to one computer.
TechCrunch frames Claude Cowork’s expansion as coding agent wars spilling into the rest of the office. The practical change is simple: tasks can run in the cloud, continue after a laptop closes and report status back to a phone.
