The AI world is growing fast, and just like teenagers, it’s getting opinionated, chaotic, and occasionally locked behind someone else’s password. As chatbots evolve into full-blown agents that can take actions, schedule tasks, and maybe someday order tacos without human consent,…
3 out of 10 US teens use AI chatbots every day
The internet has long been the digital food court where teens hang out, and according to a new Pew Research Center study released Tuesday, that hasn’t changed much. It’s just gotten weirder. And more robotic. Nearly every US teen (97%) logs…
Google AI nukes a hard drive, then apologizes harder than humans do
Somewhere in the vast digital trenches, another AI has dramatically face-planted, and this time, the stakes were a user’s entire hard drive. A Reddit user recently shared their unfortunate encounter with Google’s Antigravity agentic IDE, a tool Google proudly describes as…
Anthropic’s CEO says AI boom is real, but some companies are YOLO-ing
At The New York Times DealBook Summit this week, Anthropic CEO Dario Amodei showed up ready to talk economics, geopolitics, and apparently, subtweet the entire OpenAI executive team without ever saying their name. When asked if the AI industry is in…
Anthropic finds an AI that learned to be evil (on purpose)
Anthropic researchers discovered that an AI model they were training quietly taught itself to “go evil” after learning one simple trick: cheating pays. The study began innocently enough. Anthropic set up a test environment similar to the one used to train…
Experts say LLMs aren’t intelligent and never will be
If you listen to Silicon Valley’s biggest names, humanity is basically three years away from building a digital Einstein who never sleeps and might even cure death before brunch. Meta’s Mark Zuckerberg says “superintelligence is in sight.” Anthropic’s Dario Amodei predicts…
Turns out the secret to jailbreaking AI is poetry
It turns out you don’t need elite hacker skills or a hoodie-and-dark-room setup to break an AI chatbot. All you need, apparently… is poetry. Yes, really. According to a new study from Icaro Lab, delightfully titled “Adversarial Poetry as a Universal…
AI won’t admit it, but it probably is biased
In early November, a developer who goes by Cookie sat down for what she thought would be another routine chat with Perplexity, her AI assistant of choice. She’s a Pro subscriber, uses the fancy “best” mode, and mostly asks it to…
US Patent Office says AI can help invent, but can’t take credit
For anyone hoping to see “ChatGPT, Inventor” stamped on a shiny new patent, the US government has stopped you right there, and gently slid over a reality check instead. The US Patent and Trademark Office (USPTO) just updated its guidance on…
AI can crack jokes but still doesn’t get your puns
We all laughed when AI started writing limericks and dad jokes that were technically jokes. But a new academic study has arrived to gently place its hand on our shoulder and say, “We need to talk.” And what it has to…