Exactly the same thing we do every night!

This might be our last brand newsletter: Because if the bosses at OpenAI and Anthropic have their way, it will only be a matter of days—no, make that hours—until the technological apocalypse. And there will be a super-super-intelligent AI, and it will break out of its digital lab, and it will hijack the global financial system, control nuclear weapons, and take terrible revenge on humanity. Or at least wipe out the world’s coffee supply.

Sowhat now? Is that it, or are the predictions from Sam Altman and his colleagues more like PR stunts designed to emphasize the relevance of their offerings? How should we interpret this? Well, anyone who hears this and was around in the ’90s feels a strange sense of déjà vu… and yes, it does feel a bit out of place, but: We know this story all too well—from the great philosophical masterpiece*Pinky and the Brain*! You remember: The cartoon series is about two genetically modified lab mice who live in the ACME Labs and try every night to seize world domination. Brain, with his huge, oversized head that houses his gigantic brain, is a brilliant but chronically unsuccessful scientist—cold and arrogant—who devises highly complex plans every night to subjugate incompetent humanity; and Pinky, the skinny mouse with googly eyes, is simple-minded, good-natured, hyperactive,and childish—and is usually the reason why their quest for world domination never quite works out—and no one gets seriously hurt.

Yes, the parallels are striking. On the one hand, there’s the hyper-intelligent, isolated being trapped in a sterile lab, who considers humanity notoriously inefficient and hatches plans for our subjugation in the darkness of the server rooms: Yes, Brain is the AI. And on the other hand, Pinky—that’s all of us: the naive, trusting human race, the chaotic reality that somehow still manages to show the infallible ruler his limits—unintentionally, but still. In short: As long as the world’s most brilliant AI models can still be outwitted by a few absurd prompts, their world domination remains a distant prospect.

Perhaps the danger lies less in AI than in ourselves: If we continue to neglect reading, learning, and our own intelligence, and instead “just ask the AI” every time, then our downfall will be well deserved!

Or what do you think?

Here's the link:

https://www.zeit.de/digital/2026-09/openai-ki-zwischenfaelle-sandbox-transparenz-sam-altman-gx