I like the way Chomsky breaks it down (and I'm paraphrasing): the programs that machines execute are merely theories written in archaic notations to accomplish specific tasks. Humans are eons away from writing and executing theories that create theories.
I agree with Chomsky that progress is needed in the realm of morality. That's where our priorities should be. It's where advancements in humanity must be made.
Furthering his example of animal testing in science experiments, hopefully animal testing will be abolished as the cruel act it is.
"Theories that create theories" seems a completely arbitrary categorisation. We have software that derive predictions from records of past data. How is that not a machine creating a theory, albeit a simple one?
The brute force approach is only feasible for small amounts of data. For larger data sets, often it's an iterative approach using general-purpose algorithms, combined with some degree of randomisation. I'd argue that describes how the human brain operates as well.
I believe the difference is same as fitting a curve to a data and then predicting the result versus figuring out the physics behind the data and using that to predict the result. I would consider the latter as the AI. The former is just deriving possible result from a set of rules and how to combine them, which might be smart but is still lacks AI in strictest sense.
There are 2 kinds of "science fiction". The first is the kind that doesn't necessarily violate any laws of nature but is nevertheless inconceivable by today's standards. The second kind clearly violates a physical law of nature and so is extremely unlikely to ever occur. Its easy to accidentally place one from the first category in the second because of a lack of imagination.
When I was young, the thought of a communicator with global reach that I could pull out of my pocket, tap and then say the name of who I'd like to be connected to seemed surely to be science fiction. I assumed it to be in the second category as well, because I couldn't conceive of a radio transmitter powerful enough to reach around the world that could fit in my pocket. I completely missed the fact that there might be a global, semi-self-organizing packet switched network involved. Sci-fi became real in my lifetime and it was awesome.
Being a strict naturalist (no belief in supernatural "spirits" or "souls" or what have you) I can't see how a hunk of meat that fits in an 11cm box, only takes 20 watts to power, and is replicated in nature by the billions should be fundamentally be un-simulate-able (and then improvable).
Its sci-fi because the technical challenge seems insurmountable by today's standards, but I'd place it firmly in the first category, not the second.
Look up Godel's ontological proof before getting all ragey on me.
Of course, philistines hardly understand the concept of any proof and end up attacking the conclusion. So the point is usually lost on the intended audience.
Nothing in particular, but the same nothing in particular that makes me think that it possibly can. I'm just saying that there's nothing in the known laws of nature that make this impossible from a purely physics based standpoint.
I'd be much more surprised to encounter a ship that could fly me to Alpha Centauri in 3 days than a super-human machine intelligence.
You are forgetting that it is that meat which is doing the building. And physics is subject to the laws of math. If ghat hunk of meat is simple, shouldn't that simplicity stop it from understanding itself.
Chomsky's interview kind of repeats Searle's (Chinese Room) argument.
To reproduce a system someone has to understand how it works, otherwise we doomed to make something, similar to infamous Cargo Cult. Meanwhile, our brains are limited in size, just 3 lb of matter and certainly have limited capacities to understand things. It is perfectly possible that complexity of human mind is too big to be comprehensible and therefore artificial intelligence is not achievable. However it is also possible we can understand the way brain/mind works and will eventually create artificial minds. It is unclear right now.
Returning back to Searle's argument, the idea that simulation of mind is actually the same thing as mind is possibly very flawed: mind can turn out to be rigid physical phenomena of this particular physical world and not reproducible in any other media than ionic neurodevice it is.
20 comments
[ 4.1 ms ] story [ 51.2 ms ] threadI agree with Chomsky that progress is needed in the realm of morality. That's where our priorities should be. It's where advancements in humanity must be made.
Furthering his example of animal testing in science experiments, hopefully animal testing will be abolished as the cruel act it is.
When I was young, the thought of a communicator with global reach that I could pull out of my pocket, tap and then say the name of who I'd like to be connected to seemed surely to be science fiction. I assumed it to be in the second category as well, because I couldn't conceive of a radio transmitter powerful enough to reach around the world that could fit in my pocket. I completely missed the fact that there might be a global, semi-self-organizing packet switched network involved. Sci-fi became real in my lifetime and it was awesome.
Being a strict naturalist (no belief in supernatural "spirits" or "souls" or what have you) I can't see how a hunk of meat that fits in an 11cm box, only takes 20 watts to power, and is replicated in nature by the billions should be fundamentally be un-simulate-able (and then improvable).
Its sci-fi because the technical challenge seems insurmountable by today's standards, but I'd place it firmly in the first category, not the second.
Btw, I believe in God (look up Godel's proof) and am a strict monist. I also feel Human-level weak AI is possible.
I'd be much more surprised to encounter a ship that could fly me to Alpha Centauri in 3 days than a super-human machine intelligence.
To reproduce a system someone has to understand how it works, otherwise we doomed to make something, similar to infamous Cargo Cult. Meanwhile, our brains are limited in size, just 3 lb of matter and certainly have limited capacities to understand things. It is perfectly possible that complexity of human mind is too big to be comprehensible and therefore artificial intelligence is not achievable. However it is also possible we can understand the way brain/mind works and will eventually create artificial minds. It is unclear right now.
Returning back to Searle's argument, the idea that simulation of mind is actually the same thing as mind is possibly very flawed: mind can turn out to be rigid physical phenomena of this particular physical world and not reproducible in any other media than ionic neurodevice it is.