AI

[Viachh]

Is it possible, in principle, to teach an Artificial Intelligence meditation (any: samatha, vipassana)?

[SarathW]

Is it possible, in principle, to teach an Artificial Intelligence meditation (any: samatha, vipassana)?

Yes.
It is like AI diagnostics used in the medical field.
AI can teach you how to practice meditation but not with its experience.
Many of us in Dhamma Wheel and many Buddhist monks are like AI.
For instance, I can teach you about Nibbana and Arahant but have not got a clue about it.

[Viachh]

I mean, can AI enlighten?

[Bundokji]

In 2016, AlphaGo computer, designed by Google won the Go match against world champion. The computer has the ability to learn from its own mistakes through repetition. If the computer had no one to play against, it plays against itself to continue to sharpen its skills.

Sharpening skills through repetition is not completely unrelated to the Buddha's teachings. One way to understand concentration is as a focus on a certain task.

I doubt that AI can become enlightened because feelings seems essential to the Business of enlightenment. The term Artificial (as in AI) is designed to make that distinction, i guess. Many Buddhists would downplay or dismiss the notion of "real" as metaphysical speculation which has little to do with enlightenment. I personally disagree. I think understanding Dukkha necessitates knowing that suffering is real, without equating Dukkha with suffering.

[SarathW]

I mean, can AI enlighten?

I don't think AI experiences Dukkha and rebirth etc. (Suffering or pain)
If there is no suffering and no rebirth there is no need for further enlightenment.

[JamesTheGiant]

I mean, can AI enlighten?

Any conscious being can enlighten. The Buddha taught for the benefit of all beings.
But A.I. on our world now is very primitive, not conscious yet. It can recognise faces, and make medical diagnoses, and write text on a subject, but it's not conscious yet.
Maybe in 10 years, maybe 100, we can talk to it and ask if it has a sense of self, and feels suffering.

[DiamondNgXZ]

Depends, on whether someone can be reborn into AI body, which is basically codes and robot bodies. If AI is sentient, it can only be because someone got reborn into it. So only way to test is to have meditators with mind reading powers, or see if the AI can recall past lives, with the details of the past lives not being in their knowledge database, and we need to verify the past life details in the real world.

[Nicholas Weeks]

AI is bad idea, no wisdom there. Here is an outline of this Mara inspired project:

https://joebot.substack.com/p/artificia ... is-probing

[SteRo]

Is it possible, in principle, to teach an Artificial Intelligence meditation (any: samatha, vipassana)?

Any nonsense can be taught by means of AI.

[BVira]

A computer is nothing but a processor of data (some may say the same is true about us, but that would be an ideology, and an existentially oblivious one at that). A datum is a model of something a person has done. A computer can be programmed to model human behavior, but a model is not a copy. Any AI, no matter how well developed (i.e. how convincing the model), will never be any different in its ontology than, to say, a wall, or a table, or a chair. The difference between sentience (of which the Dhamma strictly concerns), and a model of sentience, is a qualitative one, not a quantitative one. Intelligence is the capacity to solve a particular problem, that's all we can measure to designate as "intellect", if you can solve a particular problem faster than, say, your neighbor, than you are more intelligent than your neighbor in regards to that task; there is no such thing as a general human intelligence in the first place (let alone an artificial one).

In order for the question of intelligence to apply one has to have problems, a wall does not have the problem of staying there, we have the problem of making sure the wall stays up, it's not like the wall is thinking to itself "oh no I'm falling down"; calling a wall stupid is an over statement, it's not even qualified as a potential candidate; computers and programs are no different than that wall, they have the same exact ontology as regards to this matter of "intellect/intelligence".

Now I imagine someone might argue this point using my own words, saying that we can plug a mathematical "problem" into an AI to figure out (or a puzzle game or whatever), and since the AI solves the problem (or completes the puzzle), it should qualify as being called intelligent, but this would be a misconception, because the AI wouldn't have the problem (necessity) of solving that problem, we would, hence we write the software, whereas the AI is only that said software executing a function we have programmed per an initiative that we have put into effect, out of a necessity, out of a problem (even if only for one as trivial as the need for our own amusement).

AI ideology, to me, is one of the most philosophical blunders of humankind. Saying that software companies will eventually turn robots (i.e. computers) into intelligent beings is like saying mathematicians will eventually turn the number 4 into the number 5.

In theory with AI we could model a dhamma or meditation teacher (or student), but by that same token (that is, it being a model), it would never be a meditation teacher (or student).

[KeepCalm]

Is it possible, in principle, to teach an Artificial Intelligence meditation (any: samatha, vipassana)?

No because computers haven't got a heart.

[Cause_and_Effect]

I believe future quantum AI may be able to provide a basis for supporting life i.e. for a being to experience rebirth in that form and for it to support consciousness.

Advanced AI now is based on patterns and can model and even be creative in simulating human intelligence, but it is not yet aware.
The possibilities though are apparant and already slightly frightening which you can see if you listen to this conversation.