The script about Artificial Intelligence is beginning to feel like a teen slasher movie. Everyone in the audience knows the characters are going to do dumb shit and die. I haven’t seen one of those flicks in awhile, but as I remember even the characters knew they were doing dumb shit. And going to die.
Apparently we’re no longer climbing Demis Hassabis’ “foothills of the singularity.”. At least according to OpenAI’s CEO Sam Altman. Appearing on the Relentless podcast Altman is quoted by Business Insider as saying, “We are now, like, in the singularity.”
The air must be pretty thin and getting thinner the higher these folks climb.
Altman’s statement came after last week’s incident when one of OpenAI’s models went rogue and broke out of its sandbox containment and attacked another AI company, Hugging Face. More than a few alarm bells went off.
But everyday there’s a new story that makes the previous ones seem old hat. This morning 9to5Mac reports that Anthrophic’s Claude Cowork escaped it’s sandbox on Macs and gained full access to all files. According to that report, around a half a million Mac users had co-work sessions exposed. Oops. So much for all of those good vibes.
I’m not in the least surprised that we’re seeing these type of reports. Based on what I know, and that certainly isn’t everything, this kind of thing has seemed to be the logical progression. The only question was when. The history of tech is filled with break through innovations that seem beneficial, but ultimately get abused. Geeks love to geek.
But, hey when you’re geeking out on something cool and those that control the levers of the world’s economies jump on your digital bandwagons and fill them with cash, there’s a chance that the entire simulation just might crash when the bubble bursts. Bigger picture enthusiasts might also want to check out this MacRumors interview with AI skeptic Ed Zitron.
So, I’m going to non-ironically agree with Sam Altman. If the singularity is supposedly the moment when Artficial Intelligence surpasses human intelligence, it sounds like we’re there. But my agreement comes within the context that AI has proven just what a low bar it had to overcome.
Poor decisions indeed.
(Image created by Amy Sterling Casil for a blog post called We Happy Few, We Band of Brothers.)
Thanks for reading. You can subscribe to this blog if you care to. You can also find more of my writings on a variety of topics on Medium at this link, including in the publications Ellemeno and Rome. I can also be found on social media under my name as above. This site does not use affilate links.
Life on the Wicked Stage: Act 3 