Em tempos de Claude eu digito menos mas o que mudou de verdade foi o tipo de decisão: mais trade-off de sistema, menos implementação. As vezes acho que o nome do cargo de senior software engineer já não descreve isso. A função mudou bastante pelo menos...
OpenAI DevDay 2026 dump:
dots: Grok Bot alternative
ChatGPT Space: Notion alternative
GPT-6.1 Sol: Astra-level, far cheaper
Ultrafast: 300 tok/s
Pro $500: 25x usage plan
Decisions API: almost Jev, on GPT Luna
safer computer use
Codex harness / Cloud / Security Cloud
Agents API
connect ChatGPT to third-party apps and spend the tokens already in your subscription
plugins and extensions
OpenAI Marketplace
reset
Carlos Eduardo retweeted
Meta announced Muse Charm: a keychain-sized device with a 2-inch touchscreen, a camera, at least three microphones & a fingerprint sensor to wake it. Ships December. No price.
The architecture is the interesting part.
Meta reportedly gave it built-in 5G, which means it doesn't need a paired phone & on something this size there's no room for meaningful onboard compute. So the device captures, the network carries & the model runs in Meta's cloud.
Sensor at one end, intelligence at the other.
That's the same admission as the VR Glasses at the same event, where the processor sits in a separate puck.
Both say the limit on wearables isn't chips. It's thermal load & battery mass on the body.
Zuckerberg's own framing: if you're not wearing glasses, this is the fastest way to show Muse what's going on around you. Which is a camera, on a keyring, in whatever room you're in.
The specs that actually matter haven't been published like battery life, whether capture is continuous or on demand & what signals to people nearby that it's recording. Meta says the layout isn't even finalized.
Humane's Pin ran hot & died. Rabbit's R1 underdelivered.
This one offloads both the heat & the thinking, which fixes the hardware problem but also creates a different one.
This is wild.
Claude just found something humans missed.
It spotted a hidden repeat pattern in raw bacteriophage DNA…
…and uncovered a previously unknown enzyme system in bacteriophages with CRISPR-like repeats. 🤯
950 agents. 21 hours. 210M tokens.
Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages. Beside the enzyme’s gene sits a long array of repeating DNA—a structure that looks somewhat similar to CRISPR.
We don’t yet understand what this system does, but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA. Historically, the discovery of such programmable systems has helped revolutionize medicine. CRISPR, for instance, is now the foundation of genetic medicines. But it will take much more work to learn what this system does, and whether it can be put to similar use.
Read more: anthropic.com/news/claude-di…
CCA-F
The Claude Certified Architect: Foundations is Anthropic’s first official technical credential. It sits at the entry point of a multi-level certification stack, with additional credentials for sellers, developers, and advanced architects planned for later in 2026.
Carlos Eduardo retweeted
Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages. Beside the enzyme’s gene sits a long array of repeating DNA—a structure that looks somewhat similar to CRISPR.
We don’t yet understand what this system does, but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA. Historically, the discovery of such programmable systems has helped revolutionize medicine. CRISPR, for instance, is now the foundation of genetic medicines. But it will take much more work to learn what this system does, and whether it can be put to similar use.
Read more: anthropic.com/news/claude-di…
Haiku 5.5 leaks are getting loud.
Anthropic already confirmed Sonnet 5.5 and Haiku 5.5 land in the next few weeks. Word is both are done and just waiting on the switch.
The spicy part: Haiku 5.5 is rumored to sit near Opus 5.5, maybe even press Fable 5.1 on some work 🤔 at a much lower price. Fast, cheap, safer, and closer to frontier than a Haiku has any right to be.
If that holds, Anthropic’s value model just became the one to watch.
Carlos Eduardo retweeted
BREAKING:
Anthropic researcher Jacob Coxon resigns, warning that AI systems will soon be able to hack anything and that the company is racing toward superintelligence that could kill us all by the end of the decade.
Carlos Eduardo retweeted
🚨 ⚠️ Pesquisador da Anthropic pede demissão e alerta que a corrida para desenvolver uma superinteligência capaz de se aprimorar sozinha pode destruir a humanidade até o fim da década; ele, que também já deixou a OpenAI, afirma que ambas as empresas estão “apostando” com o futuro da humanidade.
🤖 OpenAI released Astra, describing it as its most powerful and capable AI model yet, particularly for computer and browser-based tasks. Deployment begins with Daybreak cybersecurity customers before expanding to Pro, Plus, Enterprise, Business and API users. President Greg Brockman called Astra OpenAI’s most intelligent and best-aligned model, while company benchmarks show it outperforming OpenAI’s own Sol model and Anthropic’s Fable on tasks including bug detection and codebase analysis. Astra’s use of “opaque recurrence,” however, has renewed concerns about the ability of researchers to inspect AI reasoning, with chief scientist Jakub Pachocki acknowledging that monitorability becomes more difficult as capabilities increase. Brockman has gone considerably further, saying he personally believes Astra has reached AGI — although exactly what qualifies as AGI remains, conveniently enough, open to interpretation.
🤖 Google DeepMind unveiled WeatherNext 3, its most accurate AI weather forecasting model to date, which will now feed forecasts directly into Search, Google Maps and Gemini. WeatherNext 3 outperformed systems from Microsoft, Nvidia and the European Centre for Medium-Range Weather Forecasting on Operational WeatherBench, delivering forecasts at 5 km resolution, improving rain prediction by 60% and generating hourly forecasts using real-time satellite data. Google describes it as the first high-resolution global weather model to directly ingest raw satellite observations, although rival WindBorne has employed similar raw-data techniques since late 2025.