Model-Free, Model-Based, and General Intelligence [pdf]

It seems the fundamental problem with bottom up&#x2F;learning AI is that it is opaque and essentially unknowable. I find it all very hackish. We can develop systems now which we can test and seem to work, but we don&#x27;t know exactly why they work (eg: what parts of the training data they are promoting) and when (or why) they will fail. The effectiveness of adversarial inputs to trained vision systems illustrates this.<p>Zoom forward to a super-human AI that mimics our brains in its approach but exceeds its capacity. What is stopping it, for instance, learning that it can play the long game of being good until it has sufficient power at its disposal and then becoming evil? No matter what training data you present, you can&#x27;t know exactly what the result will be.<p>I get the feeling that learning systems will be combined with model systems with the former performing &quot;low level&quot; tasks and the latter providing a verifiable &quot;executive&quot; that guides high level goals or outcomes.

I attended this talk at IJCAI, and I must say that the whole system 1 &#x2F; system 2 analogy rubbed me the wrong way.<p>A solver for e.g. 3-SAT is general only in a very narrow sense, namely that an entire class of problems can be reduced to the specific problem it solves. However, the solver itself is not doing the reducing, rather it is being spoon-fed instances generated by somebody, and that somebody is doing all the hard work of actually thinking. The solver is just doing a series of dumb steps very quickly, with lots of heuristics thrown in. How is that not also &quot;system 1&quot;?<p>Anyway, the whole thing was just a fancy way of saying that you can either solve problems exactly, in the way that complexity theorists and algorithm designers do things, or statistically, in the way that learning theorists do things. No need to superimpose a strained analogy.

AGI will be a reflection of ourselves. First we must resolve the basic problems of the human condition (poverty, hunger, housing, war, ...) before developing AGI as it will surely amplify our worst nature as well as our best nature.

Was this an invited talk?

&quot;If we want good AI, we can’t look away from culture and politics.&quot;<p>At the end AI will join our tense political atmosphere of parties fighting for ruling the world?