This is a very simple proof that AGI is theoretically possible. This is not meant to be a realistic way to build AGI, just a proof that it could be done in theory if we had huge amounts of computing power. We know that:
- Humans are intelligent creatures.
- Humans are built out of atoms.
- Atoms obey the laws of quantum electrodynamics, which is Turing-computable (using a constant like pi or e to generate the random numbers necessary).
Therefore, we know that you can build an intelligent system which obeys the laws of quantum electrodynamics (because humans are such a system), and can therefore be simulated. Since this simulation duplicates every feature of a human down to the quantum level, we know that it can solve any problem a human can solve. So we know that there is at least one artificial computer system which has general intelligence at least as powerful as a human’s. QED.
No spelling errors found
Damn spellchecker sabotaged me!
What power does it take for the RE language algorithms to work efficiently as well as analyze every thought that gets processed (thoughts don’t end). The Singularity can not have a limit to what it can process and when, and especially why. Sometimes I feel like a single thought in a single nanosecond is equivalent to terabytes of stored information. We definitely need immense storage and processing capability before we can hope for the Singularity
“Sometimes I feel like a single thought in a single nanosecond”
Thoughts do not last one nanosecond, or even one millisecond; indeed, most of the uses of “nano” in everyday life are simply gross exaggerations. Neurons can only fire at a maximum rate of 200 Hz or so.
“is equivalent to terabytes of stored information.”
The entire brain does draw on terabytes of stored info; so does Google (Google has several petabytes of disk storage).
I guess Penrose would deny the “atoms obey QED” part.
And Searle, I think, would deny the part that “if a machine can solve all problems a human can, it’s intelligent”. And religious people would deny the part that “people are made of atoms”.
“I guess Penrose would deny the “atoms obey QED” part.”
QED is one of the most well-confirmed theories in all of history- it’s been shown to be accurate to twelve decimal places. See http://en.wikipedia.org/wiki/Precision_tests_of_QED.
“And Searle, I think, would deny the part that “if a machine can solve all problems a human can, it’s intelligent”.”
Then how is he defining “intelligence”?
“And religious people would deny the part that “people are made of atoms”.”
It’s just a matter of time before we get brain scanners sophisticated enough to demonstrate conclusively that the brain really does obey the normal laws of physics.
Penrose’s view is that quantum gravity effects in the microtubules in our brain somehow play a role in wave function collapse and that this has something to do with our consciousness and mathematical insight: http://en.wikipedia.org/wiki/Orch-OR . I guess that means that in some circumstances only, he would say QED with random collapse is no longer the extremely good approximation that it is everywhere else. I don’t know a lot about his theory, though.
I suppose Searle would mumble something about how the physical properties of the brain support “intentionality” and the physical properties of binary computers don’t. He’d say computers can’t actually understand things, they can just behave as if they do.
Honest religious people would have to predict that when we do get that good at scanning brains, we’ll find that there’s a non-material layer (“the soul”) on top of the brain that affects atoms but isn’t made of atoms. You can’t cite future experiments as evidence for your claim.
I think all this is obvious hooey too, by the way. I’m just pointing out where disbelievers in AI might disagree with your assumptions, however wrongly.
“Penrose’s view is that quantum gravity effects in the microtubules in our brain somehow play a role in wave function collapse and that this has something to do with our consciousness and mathematical insight:”
Wave function collapse is a mathematical event. If we can describe it with math, we can in principle put it on a computer.
“I suppose Searle would mumble something about how the physical properties of the brain support “intentionality” and the physical properties of binary computers don’t.”
This would be a very interesting debate, as computers and brains are both made of the exact same building blocks.
“You can’t cite future experiments as evidence for your claim.”
I am reasonably confident that there are no violations of QED inside the brain; I have studied enough biology to realize that we have very accurate models of how neurons work.
Penrose would deny that the collapse occurring in the brain is algorithmically computable. Of course, what this means is hard to tell, since it has to involve a function from a finite set of inputs to a finite set of outputs, and all such functions are trivially computable (lookup table).
Regarding Penrose’s theories about quantum effects in microtubules in the brain:
this whole theory of consciousness he has is part of a colaboration between him and Stuart Hameroff (anesthesiologist and the director of the Center for Consciousness Studies at the University of Arizona)
At the Beyond Belief Conference (mostly attended by scientists promoting atheism; go to “Session 4″ at http://beyondbelief2006.org/watch/) he presented their theories, including some really weird sounding ideas about moral absolutes in parallel universes that we have access to through quantum effects…
The other scientists there practically tore him apart – the humanities people basically said “I don’t understand a word of what you just said” and the hard sciences people (especially physicist Larry Krauss) told him his ideas of quantum effects in the brain are ridiculous.
I can really recommend it – his talk really isn’t all that interesting (except that his and Penrose’s ideas sound VERY strange) but the critique afterwards is a joy to watch
Pingback: Life, the Universe, and Everything » Sciences/Humanities Dichotomy
A question to religious people who believe the brain is not made of atoms:
You eat and drink atoms. Where do those atoms transform into non-material stuff that supports this non-material existence you are a proponent of?