Updated 15-Oct-2026
I like AI. It is neat, and I especially like the kind I can run on my relatively underpowered integrated-graphics mini PC at home. Yes, slow, but good enough to get actual work done. Here are a slew of things:
- Translate books
- Create audiobooks out of ebooks
- Create music
- Be a personal assistant
- Help develop software
That said, if the threats from AI are real. And they clearly are when the risks and order of magnitude of potential damage is high enough for the biggest AI creators to cry for government regulation. The lawyers are getting a bit too nervous, and are crying to to limit their potentially enormous liabilities.
If this is all true: risk is high, impact is high, then we need more than regulation, we need a plan B. Basically, if we are going to be stuck in a human zoo of the AIs making, then we need to have smuggled in various goods and activities that will make it palatable. Or in the case of simple destruction, we need to have the tools, resources, and plans to do the rebuild or reboot, whatever is needed.
The amount of resources put into this plan - plan b for humanity, should be equivalent to the likelihood of it happening, and the amount of resources needed if indeed it happens. Call it humanity assurance. Obviously this should come directly out of those organizations creating, aiding, and abetting the AI in the first place.
What we need specifically is how to deal with the potential problem of all networked electronic devices being rendered inoperable.