4 ms·
This is the "AI in a box" argument. It's highly plausible that an AGI thus confined will find a way to escape.
by kragen 5y ago
This is the "AI in a box" argument. It's highly plausible that an AGI thus confined will find a way to escape.
- fennecfoxen 5y agoI’d worry a lot more first about a system whose leaders decide to eliminate the people, or keep them as pets, while installing many of the more efficient, compliant AIs and robots, while the AI is still comfortably inside one or more boxes. At that point does it really matter whether the tyrant who controls you is human or AI?
- kragen 5y agoProbably what you would worry more about, in a situation none of us can begin to imagine, is not a useful guide to the actual risks.