4 ms·

For "an astronaut riding a horse" the system is filtering/selecting but nowhere does it understand (or claim to understand) horses or astronauts. It's giving yo

For "an astronaut riding a horse" the system is filtering/selecting but nowhere does it understand (or claim to understand) horses or astronauts. It's giving you an image that "syntactically" agrees with other images that have been tagged horse/riding/astronaut. The amazing bit is that we are happy to accept the image. Look closely at such images - they're always "wrong" in subtle but important ways, but we're happy to ignore that when we interpret the image.

dwringer 4y ago
To be fair I'm not sure I really understand horses or astronauts myself.
- dekhn 4y ago
  they're just points in a large vector space imagined by your mind for the purpose of hallucinations
metiscus 4y ago
I suspect that the issue arises from the difference in specificity about the desired result. When we say "astronaut riding a horse" we may have preconceptions but any astronaut riding any horse will likely be acceptable while asking for a specific proof of a result in mathematics has only a very few and very specific solutions. Effectively it is like the concept in math where the area of even a large number of points is effectively zero while even small polygons or regions is nonzero. Specific things like proofs are point like knowledge while the picture of an astronaut riding a horse is a surface.
halpmeh 4y ago
The situation you describe is exactly the "Chinese room" argument. I don't want to get too far into the weeds here, but the DALLE / Stable Diffusion models are cool because they do what you ask, even if they do so imperfectly. This model from Facebook cannot accurately answer a single thing I've asked it.
- visarga 4y ago
  I often hear the claim "AI does not really understand" but when you can ask it to draw an armchair in the shape of an avocado or an astronaut riding a horse on the Moon, and it does it (!!?), it's not like the "Chinese room" had any specific rules on the books on these questions. What more do people want to be convinced? AIs make many mistakes, humans make many mistakes. AIs fail to draw hands, humans too, with few exceptions. Ask a human to draw a bicycle from memory. https://road.cc/content/blog/90885-science-cycology-can-you-draw-bicycle https://road.cc/content/blog/90885-science-cycology-can-you-... Some of us believe in conspiracies rather than science. But as a species we still think ourselves intelligent.

dwringer 4y ago

To be fair I'm not sure I really understand horses or astronauts myself.

dekhn 4y ago
they're just points in a large vector space imagined by your mind for the purpose of hallucinations

metiscus 4y ago

I suspect that the issue arises from the difference in specificity about the desired result. When we say "astronaut riding a horse" we may have preconceptions but any astronaut riding any horse will likely be acceptable while asking for a specific proof of a result in mathematics has only a very few and very specific solutions. Effectively it is like the concept in math where the area of even a large number of points is effectively zero while even small polygons or regions is nonzero. Specific things like proofs are point like knowledge while the picture of an astronaut riding a horse is a surface.

halpmeh 4y ago

The situation you describe is exactly the "Chinese room" argument. I don't want to get too far into the weeds here, but the DALLE / Stable Diffusion models are cool because they do what you ask, even if they do so imperfectly. This model from Facebook cannot accurately answer a single thing I've asked it.

visarga 4y ago
I often hear the claim "AI does not really understand" but when you can ask it to draw an armchair in the shape of an avocado or an astronaut riding a horse on the Moon, and it does it (!!?), it's not like the "Chinese room" had any specific rules on the books on these questions. What more do people want to be convinced? AIs make many mistakes, humans make many mistakes. AIs fail to draw hands, humans too, with few exceptions. Ask a human to draw a bicycle from memory. https://road.cc/content/blog/90885-science-cycology-can-you-draw-bicycle https://road.cc/content/blog/90885-science-cycology-can-you-... Some of us believe in conspiracies rather than science. But as a species we still think ourselves intelligent.