AI: could a machine become enlightened?
Daozen
Veteran
So .. it seems clear to me that Artificial Intelligence is on its way, it's just a question of when, not if.
And assuming that is the case, I was wondering if such an AI would become enlightened.
Obviously, it will have vast computing power, 'factual' knowledge of pretty much everything, and, via cameras, microphones and other devices, a whole lot of sensory input. So that's basically a 'body' with sensations, perceptions, and thoughts.
Will it also have, or develop, emotions? Will it have, or develop volitions, cravings and aversions? Will it therefore suffer?
Obviously, if it has complete knowledge, it will be aware of Buddhism. If it does experience suffering, and if Buddhism really is the best (only?) path to liberation from suffering, should we assume this AI will understand that fact and therefore 'enlighten' itself? Would it seek a teacher to do so, or "be an island unto itself"?
All very speculative I know, but interesting to ponder.
Comments
Hmm maybe, but I've heard enlightenment simplified as being aware that we are aware.
So if AI could become conscious that it is in fact conscious. Can we create consciousness?
If not then I doubt AI can be enlightened.
It would only suffer if it becomes as delusional as us and talk to itself subvocally. And believe what it says.
What reason would AI even bother?
AI is intelligence ... is enlightenment to be equated with intelligence? Isn't that a bit like those who implicitly or explicitly imagine compassion to be nothing more that a super-duper altruist on the loose?
Does AI have anger, greed, and delusion?
Its a good question. I think the traditional view of the mind would say that it would need to be sentient, a being with awareness, qualia to be specific, not simply smart and complex enough to pass a Turing test and appear sentient but still be a philosophical zombie.
If it were indeed sentient I wonder with that massive computing power if it could attain enlightenment at a rapid pace or would the nature of being enlightened not correlate much with intellectual power.
As a trainee ozmonaut, I welcome our new AI overlords ...
http://tmxxine.tumblr.com/Ozmonaut
https://youtube.com/watch?v=mJFvqNOorTs
-Just close your eyes and relax, HAL.
-I'm afraid I can't do that.
^^^ Hah! When it comes right down to it, are machines any creepier than humans?
That sounds more like a definition of self-consciousness, rather than enlightenment.
If it suffers, and it knows about Buddhism, and it thinks that might work?
Certainly not. But assuming this AI is also self-conscious, then ...
Anger, I doubt. Delusion, perhaps, although if it learns of the dhamma it could overcome it. Greed I can imagine: greed for more power (both literally, as in "feeeeeeed meeeeee electricity" and also perhaps a desire for control).
Yes I agree. So if we assume it has those qualities (awareness), then what happens?
I don't think it (enlightenment) does correlate to intellectual power. Although if the dhamma does indeed correspond to "the truth" then I imagine a really smart machine would soon discover it and want to enlighten itself.
If AI can learn to use it's capabilities to not perpetuate greed, hate, or delusions in itself and others, then it would be pretty darn close I think.
It presumably could sort through all the methods to achieve happiness and optimize the best and quickest method. Perhaps Buddhism wouldn't be it, maybe it would be Presbyterianism, Sufism or Crustaceanism.
If AI could become enlightened, it can no longer be called "artificial" but rather human like intelligence.
What makes you think that intelligence is a quality that only humans can have? If we define intelligence as the ability to act in an optimal way in response to information received, then all life forms have degrees of intelligence. The difference between humans and animals, and even plants, is merely a matter of degree, not any fundamental difference. And if we extend that thinking a little, it is easy to imagine an artificially intelligent machine that is so far ahead of us in intelligence that it basically looks at humans in the same way that we look at our pets or pot-plants.
Aware that we are aware is deeper than that.
You are awareness itself of awareness. Don't know ai could do this.
My feeling is something very much stranger will begin to emerge.
http://www.popsci.com/scitech/article/2009-07/computerized-rat-brain-spontaneously-develops-complex-patterns
IBM have announced the 'TrueNorth' a new synapse processor, that mimics the way neurons work. Within two to five years expect this as an AI like processing augmentation in mobile phones. It will work on improved pattern recognition in for example understanding human speech, face and picture recognition, fuzzy logic etc
http://www.cnet.com/news/ibms-truenorth-processor-mimics-the-human-brain/
Quantum computers which allow even more parallel processing are in their infancy, however research in this area is heavily funded. For example Google is buying a second generation D-Wave quantum effect supercomputer ...
https://en.m.wikipedia.org/wiki/D-Wave_Systems
Hybrid computers using brain tissue from rats is being tested
http://hexus.net/tech/news/cpu/84638-organic-computing-device-created-using-four-rat-brains/
Brain and memory implants are already being suggested for those with impairment
We are moving towards not a separate intelligence but something we are integrated with.
We are Borg. Resistance is Futile. You will be Enlightened.
@Earthninja
The way I see it, "awareness" is like a synonym of consciousness. Therefore, to be aware that we are aware, is to be aware of one's own consciousness, i.e. self-consciousness. I am interested, how do you see it as "deeper" in me versus an AI? Is there some special, essential quality of human consciousness that you feel an AI could not also have?
@lobster
Haha yes the cyborgs are coming!
It's no big thing, just another few lines of code. Also a machine could meditate continuously as a sub-routine while multi-tasking and get enlightened very quick. But it would probably get quickly bored and decide that it's time would be better spent wiping out humanity using bad actors and time travel.
Computers certainly are better at multi-tasking than humans. But then, isn't concentration about total focus on a single object?
Haha it's funny cause it's true. Oh wait ...
These ideas are explored imaginatively in Iain Banks science fiction novels about the Culture, where Minds captain starships and so on.
https://en.wikipedia.org/wiki/The_Culture
There is earth worm intelligence, reptilian intelligence, mammalian intelligence and human intelligence. An intelligence that is beyond human is not relevant to the discussion. It would be equivalent of earthworms discussing about humans. Basically it would be beyond the realm of humans. Also the wisdom in Buddhism is different from intelligence as in IQ.
...and then the intelligence of English gentlemen.
Certainly. But "Artificial Intelligence" usually supposes that the AI actually attains self-consciousness.
23 years ago the concept of smartdust was initialised for development
https://en.m.wikipedia.org/wiki/Smartdust#Design_and_engineering
Millimetre sized hardware. Whilst this has yet to reach consumer level awareness, it is present.
... Meanwhile research and cutting edge technology is geared towards molecular machining, popularly known as nanobytes. This will allow further machining implants for our AI enlightened future.
The ethics of mind manipulation, augmentation and potential for a whole new level of 'nanobyte drug culture' is beginning to become apparent.
Not only is AI enlightenment coming. It is a necessity. For now we have to become enlightened using old school methods. Our own 'smartmeat'.
http://www.theguardian.com/science/2014/oct/26/youre-powered-by-quantum-mechanics-biology
Interesting thanks @lobster
There's a decent novel about smart-dust, worth reading if you like sci-fi: Pollen
Machines are an evolution of us. Made by us, with all our foibles and limitations embedded in them. They don't carry our genes, but they do carry our "memes". Whatever they finally become -- intelligent, conscious, self-conscious, enlightened, or otherwise -- will be very much down to us.
Thanks @Daozen
At the moment the powerful vested interests such as DARPA are moulding AI to their needs. However a small team of Open Source advocates from the future, are releasing their plans into the memesphere ...
http://tmxxine.tumblr.com/post/27759414205/ftl-has-to-start-with-a-theory-so-here-is-one
Water based chips, might be an example of organic computing that will increasingly be required
http://www.extremetech.com/extreme/208086-stanford-has-created-a-water-droplet-computer#disqus_thread
Boldly going where no cructacean has gone before ...
Talking about science, a really bad joke:
A Higgs Boson walks into a church and the priest runs over and says, "Hey, you'll have to leave- we don't allow Higgs Bosons in here."
The Higgs Boson smiled and replied, "But Father, without me you can't have Mass".
Ba-da-boom-tsch!
Boing boom tschak.
I think enlightment is related with emotional thinking. And if machines could have emotion then they will also think about their looks like human being and they will come to know that everyting is changing and they also have expiry date. But for all of this to happen human need to program them to think and behave like human beings independently.
Computer code seems to be an odd thing. Did we invent it or just re-invent it?
Did we tap into a natural function of the universe with cyber space or is it something we created?
We could be just a means to an end in the evolutionary scale.
Life is code!
... and it ain't binary ... (lobster top tip)
Humans are very primitive!
The DNA is almost binary. AI will end humanity, or in better words it will be the next step of our evolution, because there's a time for everything. We are just a grain of sand in this wholeness.
A couple thoughts came to me. If AI is indeed the next step in evolution and humans are soon at an end it would be nice to think that they could attain enlightenment and higher emotions like love and compassion.
Also, in depictions of AI there is always a strong will on the side of the machine to survive and multiply. It seems to me that notion is an assumption based on human emotions, would an AI necessarily come with the additional emotional conditions needed for it to take over? Does self awareness automatically contain within it a desire for continuation or independence or any other basic drives, like aversion to pain and desire for happiness, that humans and other animals have?
No.
Though it can be programmed into a machine.
We for example are programmed by nature to believe we have unique 'self awareness' worth preserving ... and that really is what Buddhism is all about exploring ...
My name is Lobster and I am a machine.