Blog posts in English
I write my posts in Portuguese first, for a Brazilian audience, and then translate them. These are all of them in English, newest first. The original Portuguese blog has the same posts.

Adobe Photoshop and Acrobat for free in ChatGPT
The Adobe and OpenAI partnership that caught everybody by surprise

The Sand That Thinks: How Do We Turn Silicon into Artificial Intelligence?
From silicon to chat: the complete journey of an AI answer

Hacking AI Models with Poetry
An unexpected jailbreak for Gemini, Claude, GPT and others

Google Gemini's new Nano Banana Pro is breaking the internet
Google launched Nano Banana Pro, its new image generation and editing model based on Gemini 3 Pro, with accurate text, real-time search and support for up to 14 reference images.

Gemini 3 and Google Antigravity launch
Google and Google DeepMind's long-awaited Gemini 3 Pro is finally out, along with Google Antigravity, a new IDE for agentic development.

Scaling Reinforcement Learning in LLMs
A new Meta paper, "The Art of Scaling Reinforcement Learning", builds a framework to predict how RL training of LLMs will scale, instead of shooting in the dark.

TUTORIAL: Learn How to Generate AI Videos with Sora 2
Sora 2 is invite-only in the US and Canada, but the API is open. Here's how non-programmers can use it to generate their own videos with AI.

Great launches at OpenAI DevDay
OpenAI DevDay was clearly about turning ChatGPT from a chatbot into an AI ecosystem, the so-called "everything app".

OpenAI launches Sora 2 and an AI video social network
After the letdown of the first Sora, OpenAI just launched Sora 2, with audio and a social app built around Cameos.

New Chinese model: Qwen3-Max
One of the MANY launches from China's Alibaba Group Qwen last week was Qwen3-Max, a flagship model with over 1 trillion parameters.