[originally posted at https://www.lesswrong.com/posts/ALsuxpdqeTXwgEJeZ/could-a-superintelligence-deduce-general-relativity-from-a] Introduction: In the article/short story “That Alien Message”, Yudkowsky writes the following passage, as part of a general point about how powerful super-intelligences could be:
Great article, I think it's even worse than you presented here.
> Take two apples that are 0.1 kg each [...] The AI could figure out that the explanation “masses attract each other” is possible, of course.
Even assuming it conceptualizes mass, how could it deduce the amount from a couple frames? The apple could be like a helium balloon, heck it could be like anti-matter which until recently some scientists thought might not fall down.
> Simulation theories: [...] Any laws of physics or that are possible in a simulation are possible physics that the AI has to consider.
You know what's worse for this deduction than simulation theories? Art theories. I might just be showing the AI an animation I made which could be following no physics at all.
Great article, I think it's even worse than you presented here.
> Take two apples that are 0.1 kg each [...] The AI could figure out that the explanation “masses attract each other” is possible, of course.
Even assuming it conceptualizes mass, how could it deduce the amount from a couple frames? The apple could be like a helium balloon, heck it could be like anti-matter which until recently some scientists thought might not fall down.
> Simulation theories: [...] Any laws of physics or that are possible in a simulation are possible physics that the AI has to consider.
You know what's worse for this deduction than simulation theories? Art theories. I might just be showing the AI an animation I made which could be following no physics at all.