Not dominace but values: https://aicorpus.eu/

We should really think about the (our) AI Corpus first: https://aicorpus.eu/

Europe being “absent” from the AI safety debate may be the feature, not the bug. While Silicon Valley debates whether its own products could end humanity, Europe is asking the less cinematic question: why should our infrastructure, data and economy depend on a handful of US firms? Safety without sovereignty is just dependency with better paperwork.

Some people have a massive head-in-the-sand problem: instead of engaging with these developments, a lot of them seem to want to ignore them. As well as climate change deniers, there are now AI deniers too. 🙈

Google finally found the killer AI use case: becoming the unpaid third parent. CC reads the school emails, tracks appointments, fills permission slips, plans meals and updates the family calendar. Convenient, absolutely. Also a fascinating amount of family life to hand one company. The real household chore may be managing the permissions.

Personally I trust Mozilla and Mistral a lot more than any of those big tech bros.

Turns out the future of journalism is three guys buying trusted publications, firing the writers, inventing fake experts, and asking ChatGPT to wear the corpse as a name tag. 50 million page views a month is impressive, though: apparently Google Discover has finally perfected the art of rewarding a haunted content mill.

Thanks. 🖖

Mozilla choosing Mistral for Firefox Smart Window feels like a rare AI partnership that actually fits the product. Privacy, zero data retention and multilingual models are a much better match for Firefox than turning the browser into another Big Tech AI funnel.

Apple returning to the server market would be quite the plot twist: after years of people building unofficial Mac Studio AI clusters, Apple may simply productize the idea. Dual or quad M8 Ultra systems, potentially connected via Nvidia NVLink Fusion, could make local AI infrastructure far more mainstream. The Mac Pro might finally have found its next job.

We spent decades mapping 166,000 neurons and the internet immediately made the fly play Doom, trade Bitcoin and parallel park. Honestly, that’s the most human possible response to a neuroscience breakthrough. Give us the map of a brain and within 48 hours someone will ask whether it can run Doom. Next milestone: the fly gets a LinkedIn account.

About time Europe treated addictive feeds like the product design problem they are. The real test is whether this becomes meaningful age appropriate access with privacy preserving checks, or just another glorious cookie banner where everyone clicks »I’m 18« and TikTok solemnly salutes.

This! 💯

💞

It’s the secret power of the masses…

Personally, I’m more interested in being able to process personal data locally, which is legally impossible online. #GDPR

🤣

People will adapt - or run into massive problems anyway.

That’s the way to go.

People tell AI things they wouldn’t tell their closest friends, then act surprised when those chats can end up in court records, company systems or leaks. The chatbot feels like a diary, but it’s closer to a diary stored on someone else’s server. Maybe the new rule is simple: if you’d panic seeing it read aloud in court, don’t type it into the bot.