Some people are trying to push these AI systems in the direction of being autonomous. There's no evidence yet that auto GPT can actually do this stuff. But there is one project called chaos GPT where someone basically used auto GPT and gave it the explicit instruction to try to destroy humanity. That should probably be illegal, right?
Aric Toler untangles the web of teens, gamers and memes at the heart of the latest intelligence scandal.
Then, an update on Twitter — where things have gone from bad to worse.
Plus: How A.I. is bringing us closer to “Westworld.”
On today’s episode:
Aric Toler is the director of research and training at Bellingcat, the Dutch investigative site. He worked with journalists at The New York Times to identify the man who allegedly leaked top secret documents on Discord, a social media chat platform.
Twitter took steps to block Substack newsletters from circulating on the platform after Substack, a publishing start-up, started a Twitter-like service.