by 
Hello! This week our stories are focused on security and storms, both figuratively and literally! This week on the Brief we discuss the red herring from the OpenAI security breach, why Uber is letting you hijack drivers’ cameras, and the preposterous partnership between the Weather Company and…Kalshi?! You can bet this is a jam-packed episode.
These AI breaches are dark, ominous, and complicated, and the threat of a truly destructive lightning strike is being exacerbated by one thing: speed. The velocity at which Big Tech continues their strategy of “ready, fire, aim’ is blistering and akin to jumping into a swimming pool with full armor and two metal swords during a thunderstorm…you’re just asking to get struck.
That’s exactly what these security breaches feel like. Frontier AI Labs like OpenAI and Anthropic are basically goading their own technology to break the rules, because in some perverse way, they believe it’s a badge of honor to have built a model “this powerful.” I go back to Anthropic’s blog post on their breach a few weeks back, which read like a petty, 5th grade, “anything-you-can-break-I-can-break-better” explainer. When cyber security and crime is what you’re comparing, maybe mix in a little humility!
So, who has the answer to what’s next? Should we fire up Disney+ and watch the X-Files episode “Kill Switch” from 1998 (which is an eerily accurate extrapolation of what AI can do)? No? Well, yes, you should watch The X-Files, it’s a fantastic show and ahead of its time, but no, we’re not going to be able to ask Mulder and Scully to uncover something as simple as a “Kill Switch” to stop today’s madness. We must keep calm and carry on, don’t anthropomorphize AI, and stay abreast of what’s happening. The last part is the hardest, but that’s what I’m here for.