You'll find pretty quickly that the exact techniques that worked so well in learning to play Atari games very often fail spectacularly when you have to introduce goal-steering.

Reinforcement learning turns out to be fantastically clever at finding really stupid solutions if you give it the tiniest opening to do so. You put a camera in a room to provide feedback for an agent to learn to move an object closer to the camera, and it will happily learn to knock the camera over.

To be fair to our AI brethren, this is true of Humans too. I watch people game metrics every single day, and many of them fail spectacularly when presented with real world problems outside of their experience.

Gaming metrics is the very soul of corporate life

Then which is real and which illusion?

Good question. Sometimes, the metrics must be gamed to make real progress. (As opposed to the more usual gaming them for personal gain.)

Between these two, some self serving bias and delusion, mission statements etc.. what is real?!

If they meet the spec, they're not stupid solutions, it's a stupid spec.

The problem is that we generally rely on human intelligence to fill in gaps in specs that are really not stupid for a human. AI will exploit gaps that a human would basically say aren't there until they see evidence to the contrary.

So we can say that's a bad spec if we like, but what's the answer that leads to? I don't need an AI if I can just declare all users must be really good programmers spending their days writing unambiguous specs. That wasn't really the goal of the AI though.

Unfortunately, when gaming metrics has real world consequences (and this is of course the whole point of those metrics, assuming they're not gamed), then things like 2008 happens, and tens of thousands of people lose their homes.

I think you are misusing the term "perfect information." In this case, chess is a real world game where both players have perfect information. That is, neither player can keep a secret of where any of the pieces have moved.

So, the complexities of how the AI is taught to interact ultimately don't matter. It may have a lot of effort to parse the visual of the board to get the perfect information, but the game is defined as one of perfect information.

Wikipedia calls out that there does not seem to be consensus on what this term means for games with chance or concurrent play. https://en.wikipedia.org/wiki/Perfect_information

True. I think my point was more about games with non-trivial rules (or many degrees of freedom). For example going from chess/go to turn-based video games like civ, to starcraft. Usually it involves vastly more possibly positions in time and space.

Even then, it still reduces to a brute-forceable game with finite, explorable states; 'just' with an extra layer of (granted, quite interesting and technically impressive) parsing. We don't know what the case is for reality.

And yes, we can do our best to turn real life in to a game, but all such models leak pretty badly, and the leaks tend to cause much more fundamental instability.

I think the main problem is with simulating enough games/tries.

In the game, you can easily let the AI play against itself countless times. Much, much slower in real life.

Until we have good enough simulators/approximators of reality, AI can't learn fast. However, they can already learn driving in GTA, so who knows?

> However, they can already learn driving in GTA, so who knows?

Yeah, that sounds like it'll end well.