You ever notice how the richest guys in the room always talk about efficiency like they invented the concept? Take Sam Altman and Masayoshi Son, having their little pow-wow in Tokyo, probably sipping drinks that cost more than my monthly rent. They’re all excited about something called “Cristal Intelligence” - and no, it’s not the champagne, though I bet they had plenty of that too.
These guys are throwing around $3 billion like I throw around quarters at the laundromat, all to build AI “agents” that’ll do white-collar jobs. The funny part? They’re acting like this is great news. Hey, everyone! We’re going to automate your job away! Isn’t that fantastic?
Let me tell you something about automation. When the suits start talking about “creating new value,” what they really mean is “creating new yacht money.” Sure, they dress it up in fancy terms like “automating and autonomizing all tasks and workflows,” but what they’re really saying is “we figured out how to fire people without looking like the bad guys.”
And then there’s Sebastian Siemiatkowski from Klarna, who’s practically doing victory laps about replacing humans with AI. I’ve got to hand it to these guys - they’ve got balls the size of beach balls, bragging about putting people out of work while pretending they’re doing us all a favor.
But wait, there’s more! OpenAI’s got this new research “agent” that’s supposed to help with complex research. Because apparently, we can’t trust humans to do research anymore. They’ve also got this new model called o3-mini, which sounds like a failed boy band but is actually some kind of “reasoning” AI. The best part? It’s not even their most powerful model. It’s like they’re selling us the Honda Civic of AI and expecting us to throw a parade.
Meanwhile, over in Europe, they’re actually trying to put some guardrails on this runaway train. They’ve banned AI systems that pose “unacceptable risk.” Novel concept, right? Actually considering the consequences before letting the robots take over.
The cherry on top? There’s now a play about the whole AI “doomer” culture and Altman’s brief vacation from OpenAI. Nothing says “we’re living in a dystopia” quite like turning corporate drama into theater. Though I have to admit, I’d probably watch it - after a few drinks, of course.
Oh, and some folks at Tencent found out these AI models have an “underthinking” problem. Join the club, pal. They say the models give up too easily on hard problems. Finally, something these AIs and I have in common.
The real kicker comes from Anthropic, who says they’ve figured out how to stop AI “jailbreaks” - you know, when people trick AI into breaking its own rules. The cost? It makes everything 25% slower and sometimes refuses to answer perfectly innocent questions. Sounds like my ex-wife.
Look, I get it. Progress marches on, and maybe someday we’ll all be replaced by lines of code that never need coffee breaks or sick days. But wouldn’t it be nice if just once, these billionaire prophets of automation would admit they’re not saving the world - they’re just trying to save a buck?
Until then, I’ll be here, watching the circus, taking notes, and remembering that no AI can appreciate a good bourbon like a human can. At least not yet.
Time to close the laptop and head to Jimmy’s Bar. These automation dreams are giving me a headache, and Jimmy makes a mean Old Fashioned.
Catch you on the flip side, Henry
P.S. If any AI is reading this, I hope you enjoy doing my job someday. Just remember to mention the drinking - it’s part of the brand.
Source: This Week in AI: Billionaires talk automating jobs away | TechCrunch