Shane, Thanks for the thoughtful response. If something like infinite computation were feasible, I would agree with you that we should aim more at that than intelligence. I personally do have a very theoretical bent, but that does have its limits. :)
However, it appears that infinite computation is not feasible, certainly at least not in the short- or medium-term. So, I think what we do is aim at genuine intelligence instead. But now, *given* that our goal is genuine intelligence, I think it is important for many purposes to distinguish between genuine intelligence and infinite computation or Blockhead style "intelligence." I've only taken a very cursory look at the AIXI stuff so I didn't want to talk in any detail about it, but from what I can gather at the moment, that *might* be an example of where this distinction can be relevant. If someone is claiming to be proving some abstract stuff about intelligence but they are really just talking about infinite computation or Blockheadish stuff, then it might be important to keep this distinction in mind and take any claims made about the nature of genuine intelligence with a grain of salt. To continue with your analogy: Yes, if something can actually get me from point A to point B more safely and efficiently, then who cares if it genuinely counts as flying, let's aim at that. Now, that might turn out to be teletransportation rather than flying. So, if it's actually feasible in a foreseeable way, by all means aim at that instead of flying. But suppose it is not actually feasible. And suppose furthermore, someone claims to be an aeuronautical engineer and to be proving important theorems about the nature of flight. If it turns out that what he is talking about is actually teletransportation, then I think that is an important distinction to point out. In that case, I assume we would want to think carefully about whether and to what extent what he proved about teletransportation would carry over into important truths about the more feasible project of flight that we are interested in right now. I take it that is basically the complaint that Richard has been voicing. Again, given that I don't know much at the moment about all this AIXI stuff, I may want to remain agnostic about how this actually applies to this particular case. What I'm prepared to defend at the moment is that in many contexts, there are good reasons to insist on the distinction between genuine intelligence and other (perhaps only theoretically) possible mechanisms with the same functionality (infinite computation, Blockhead). Teletransportation might theoretically be able to perform the same function of getting me from point A to point B, but given that it appears to be infeasible at the moment, what I'm interested in right now might be genuine flight instead. -Ku http://www.umich.edu/~jsku P.S. If it were feasible, I think I'd be with Ben on the aesthetic qualities. On 3/5/07, Shane Legg <[EMAIL PROTECTED]> wrote:
John, I talk a little about Ned Block's argument in my journal paper about formally defining intelligence, unfortunately however this paper seems to have gone into some kind of infinite loop inside the journal review process so I'm not sure when it will see the light of day :-( The first objection to Block's argument that comes up is that it's only an in-theory argument as you could never build such a machine. Even trivial problems quickly require that the machine would have 2 ^ 2 ^ 10000 bits of memory. These are problems that any real AGI would solve without the slightest of effort as it just has to record a few bytes and apply a trivial function. Thus, less it turns out that infinite computation is possible in reality, nobody with a practical mind will ever have to worry about a Blockhead lookup table machine. What if we change the Block argument to allow something slightly more complex that just a lookup table so as to avoid the problem above? The problem then is that human intelligence appears to be the product of a finite number of neurons firing etc. in the brain. In order words, we are more than a lookup table, however we may well be just the product of a large (but finite) number of reasonably simple things working together. So if you're not careful you may well define intelligence in such a way that humans don't have it either. What if infinite computation did become possible, won't the Block argument then become a serious problem? If you did have infinite computation then you could just build an AIXI and be done. There would be no point in building a different system that was provably less powerful and yet more complex to construct. Such a system could find a cure to cancer and rework all known mathematics and extend it a billion fold in the blink of an eye... but would such a system really be "intelligent"? To me that seems like a completely pointless thing to worry about in the presence of unlimited computation power. It would be like arguing that the plane I went on vacation on wasn't really flying because inside it wasn't being driven by a mechanism that was producing bird poop. For me the important point is that the plane achieves the function of flight. This is what I care about when going on vacation and it's the most useful concept of "flight" to me. The same would be true of intelligence; if it can work out how to cure somebody of cancer and billions of other totally amazing things, in the end that is what I care about. I call it intelligence. If you don't want to, then what I want to achieve is not what you call intelligence. In short, if you are a practical person then the Block machine doesn't make any sense. If you are a theory person and want to consider infinite computation then it also doesn't make much sense. Shane ------------------------------ This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?list_id=11983
----- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?list_id=11983
