Should AI help you get away with killing your spouse?
What does a world of total user-aligned AI actually look like? Quick question: Do you want AI to be so well trained, it could help husbands (or wives, for that matter) plan the perfect murder of their spouses? Just as a gut reaction, that feels like a no for me.
Key Takeaways
- I wouldn't even think it was a particularly hard question.
But America contains many diverse perspectives, and one such perspective was shared by Comma AI founder and longtime jailbreaker George Hotz over the weekend.
- (I agree with a lot of what he says here!)
For Hotz, the best approach to AI alignment and safety is to focus on locally controlled AI models that are closely aligned with the interests of their users.
- (I feel like there are other rules against this?)
A truly aligned AI would be able to order meth-lab equipment from Amazon Prime and show you how to use it if that's what you wanted and asked for, he says.
- requires balancing equities, binding individual needs into a network of interdependent preferences and systems of accountability.
And anyone deploying mass-market tech products should probably think about that network as a whole, which means taking seriously the interests of the as-yet-unmurdered spouses and stepparents of the world.
- He can be reached at russell.brandom@techcrunch.com or on Signal at .
I wouldn't even think it was a particularly hard question. But America contains many diverse perspectives, and one such perspective was shared by Comma AI founder and longtime jailbreaker George Hotz over the weekend. The post comes in response to a bunch of big-picture AI alignment plans, most recently the AI 2040: Plan A policy paper from the AI Futures Project.
That paper envisions a world in which the world's researchers collectively choose to slow down AI development for 14 years for the good of humanity. But of course, not everyone who read the paper agrees with its premises or conclusion. In fact, Hotz disagrees with the whole premise that AI progress should be managed for the collective good.
In his post, he argues that the fast-takeoff scenario - the hypothetical where AI rapidly obtains superhuman abilities - doesn't make a lot of sense. (I agree with a lot of what he says here!) For Hotz, the best approach to AI alignment and safety is to focus on locally controlled AI models that are closely aligned with the interests of their users.
For more details please read the original article at TechCrunch AI.
Continue Learning
Comments
Comments appear only after moderation. Your email identifies your submission to the moderator and is never displayed here.
No approved comments yet.