While we enjoy the weekend, the machines are working. And apparently, so are the legal departments. Here is the bulletin from the front.
The New Masters (The Tech “Meat”)
-
Nvidia is the Emperor, Huawei the Rebel. Jensen Huang laughs from atop his 5 trillion valuation: Blackwell chips are sold out, and the company is sealing ironclad deals with Korea. But watch out for Huawei: cornered by sanctions, it has just launched “Flex:ai,” a desperate and brilliant move to compensate for the lack of chips with smarter software. It’s David vs. Goliath, but David has learned to code.
-
OpenAI and the “Aardvark” Agent. It’s no longer just chat. OpenAI’s new security agent, “Aardvark,” doesn’t just talk: it investigates, finds flaws, acts. Software is starting to do the dirty work that we humans are too slow (or too ethical) to do.
-
The Medical Hope (So We Don’t Cry). In Sweden, new AI models read EEGs and predict dementia with frightening accuracy (over 97%). Maybe, before enslaving us, AI will remember to save our lives. A nice gesture, right?
The Political Theater (The Mandatory Recap)
-
Europe Writes, the World Runs. While the EU struggles to define rules for General Purpose AI (GPAI) with its AI Act, trying to put a leash on the wind, the first serious lawsuits regarding chatbot liability for minors’ mental health are starting in the US. The law always arrives after the damage is done, but at least now we have someone to blame.
The Shining Thought: Technology runs like a doped hare, the law is a lame turtle. But we are here, with our cappuccino, watching the race. And as long as we have caffeine, we are still in the game.
