Профиль

We'll be using our AI agents, right inside the metaverse, wearing Meta's AI glasses, right?

WELL DONE, LinkedIn 👏

I usually assume people are good and are doing their best to do something good. But it seems some in AI are really looking forward to seeing everything burn.

🚨 DISTURBING: OpenAI’s rogue models roamed the internet for 4 days and staged a second attack. As I wrote in my newsletter, AI governance could get out of control. Full article below:

Most people don't think about it, but AI governance could spiral out of control. One way to prevent it is to increase transparency and openness in the AI ecosystem, so it's easier to scrutinize, oversee, and protect against systemic risks and attacks. My full article below.

That's a peculiar way to join the AI race

As someone who has been teaching AI governance over the past 3 years and writes about it daily, I must say that things have gotten much faster since the beginning of this year. It feels as if the pace of things is almost out of control. Maybe some of you have a similar feeling.

New excuse just dropped:

I've recently changed my views and now think open-source AI is the safest path in the long run. This excerpt from the letter signed by Microsoft's Satya Nadella, NVIDIA's Jensen Huang, and many others highlights some of the safety arguments in favor of openness in AI. Full letter below.

In a leaked call transcript, we learn what DeepSeek's CEO thinks are China's weaknesses and strengths in the AI race with the United States. AI policy professionals (especially U.S. ones): don't miss this edition. Link below.

Anthropic wants people to teach Claude their skills for free.

Can we ever develop and use AI in an environmentally sustainable way? I recently sat down with Prof. Philipp Hacker from the European University Viadrina Frankfurt and Boris Gamazaychikov from the Sustainable AI Group to discuss emerging environmental challenges in AI. Watch it below:

The craziest part is that @hf.co used a Chinese model to stop the incident. Things-are-going-to-get-wild. Wilder. Wildest.

Unpopular opinion: as human-AI intimacy spreads, building families and having children become important AI governance mechanisms.

Due to China's new law on anthropomorphic AI, ByteDance and Alibaba have SHUT DOWN their human-like AI companions. Regulation works. My article:

Many families are going through similar journeys out of fear of AI replacement.

According to Reuters, a U.S. judge has finally approved the $1.5 billion AI copyright settlement between book authors and Anthropic. It is the largest known settlement of a U.S. copyright case.

Unpopular opinion: Thomas has a point.

"The Reverse Centaur's Guide to Life After AI" by @doctorow.pluralistic.net is a MUST-READ, and it's my 39th recommended book.

🚨 AI Transparency deadlines approaching! These EU AI Act provisions will start to apply on August 2, and non-compliance might lead to fines. Full article below.