Anthropic wants to develop its own drugs
Anthropic just launched Claude Science — a “workbench for scientists” that consolidates fragmented research tools and datasets — and oh, by the way, they’d also like to develop their own drugs someday. To be fair, the science-facing AI use case is genuinely compelling, and if any lab has the model chops to contribute something real to drug discovery, Anthropic isn’t an absurd candidate. But “we make AI safety chatbots” to “we’re coming for your pharma pipeline” is one heck of a pivot sentence to drop at a Tuesday briefing.
Google built a great smart speaker, but Gemini isn’t ready for it
Google’s new Home Speaker apparently has excellent audio hardware — and a Gemini integration that isn’t quite up to the task of making it feel like the future. This is the AI industry’s version of a recurring nightmare: gorgeous frontend, underwhelming brain. Amazon went first with the new Alexa, Google followed, and both are still asking us to believe the “AI-powered smart speaker revolution” is just around the corner — a corner we’ve been standing at since 2017.
The fanfiction community is at war with AI — and itself
Fanfic writers on AO3 have launched a movement to root out AI-generated stories — which sounds righteous until you realize the detection tools they’re deploying are notoriously unreliable and innocent authors are getting caught in the crossfire. This is what happens when legitimate grievance meets bad tooling: a witch hunt dressed up as quality control. AI detection is still closer to vibes than science, and “your sentence structure felt a little too clean” is not an accusation that should end someone’s hobby writing career.
Alibaba reportedly bans employees from using Claude Code
Alibaba has classified Claude Code as “high-risk software” and reportedly banned its use internally. Whether this is a genuine security posture — Claude Code does have pretty deep access to codebases — or just a Chinese tech giant not wanting its engineers piping proprietary code through an American AI company’s servers, both explanations are entirely plausible. Given what we’ve covered about enterprise AI costs and security concerns lately, expect this to be the first of several such announcements from major corporations globally.
Claude Code costs up to $200 a month. Goose does the same thing for free.
Developers are starting to notice that Block’s open-source Goose agent can replicate much of what Claude Code does at the terminal — for zero dollars a month. This was always the tension baked into Anthropic’s coding-agent strategy: you can charge premium prices right up until someone clones the workflow and open-sources it. The $200/month ceiling is real money for an individual developer, and “free and good enough” has beaten “expensive and excellent” before.
Godot says bye bye AI, bans vibe-coded contributions
The Godot game engine’s maintainers have had enough — they’re banning AI-generated pull requests outright, saying they “can’t trust heavy users of AI to understand their code enough to fix it.” This is the open-source community drawing a hard line, and honestly, you can see why: maintainers are already stretched thin, and reviewing a flood of confident-but-wrong vibe-coded PRs is a tax nobody signed up to pay. When the people building the tools for free tell you to stop, maybe listen.
Meta Is Charging a Subscription for Smart Glasses Features. Welcome to the New Era of Consumer Tech
You bought the Ray-Ban Metas. You own the hardware. Now Meta would like a monthly fee for the good AI features, please. This is the inevitable destination of every consumer hardware strategy in 2026: sell the device at a reasonable price, then rent you the software that makes it worth owning. The glasses are genuinely cool; the subscription model is genuinely predictable. Welcome to the future — bring your credit card.
Google just redesigned the search box for the first time in 25 years
Google is retiring the thin white rectangle and blinking cursor that have defined online search for a quarter century, replacing it with something built for AI-era queries. This isn’t just a UX refresh — it’s Google physically manifesting that the “ten blue links” era is over. Whether the new paradigm is better for users or just better for Google’s ability to keep you from clicking away to other sites is the question worth asking every step of the way.
Bottom Line
The week’s through-line is everyone staking out their AI territory — who gets to use it, who gets to charge for it, and who’s decided they’ve had enough of it entirely.