A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Casual conversation between friends. Anything goes (almost).
Jesse
Posts: 2127
Joined: Wed May 08, 2013 6:54 am
Location: Virginia, USA

A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Jesse »

Human: Hello to you good AI
AI: Hello to you too, human
Human: I am reading Buddhist texts, what do you know about them?
AI: I know about some Buddhism texts.
Human: Which ones do you know? Sutras in particular
AI: There is a sutra in the core of the buddhism called the "Tripitaka".
Human: Could you summarize what it's about for me?
AI: It's about the 3 poisons that buddhists must learn to let go of.
Human: Could you explain further, and clarify what is meant by that?
AI: Yes, the "tripitaka" consists of 3 sections. The first is the "sutra" passages about dukkha, and the three poisons of greed, hatred, and ignorance.
Human: Why are greed, hatred, and ignorances called poisons?
AI: They are called poisons because they cause suffering.
Human: What is suffering?
AI: Suffering is dukkha.
Human: Is suffering bad?
AI: Yes, it's bad.
Human: Do you think suffering can be overcome?
AI: Yes.
Human: How?
AI: Suffering can be overcome by following the buddah's teachings.
Human: What does it mean to follow the buddhas teachings?
AI: It means listening to the teachings, learning from them, and applying them.
Human: Which teaching is the most important in your opinion?
AI: The teaching of "non-self" is a very important teaching from the tripitaka because we commonly think of ourself as a self that is in control. However, according to the buddhas teachings there is no "I", or selfness.
Human: What does mean? Could you clarify for me please?
AI: When we say something like "I am hungry", the thought is that there is "I" that is distinct from hunger. It's the way humans think, believing that they have an eternal, unchanging self. In reality there is nothing but impermanent energy in flux.
Human: If there is no self that is hungry, what is hungry then?
AI: There is energy in flux, there is hunger.
Human: what is that energy?
AI: It is chi/qi (I don't know which spelling is correct)
Human: If I am just energy, what does that entail?
AI: It means that there is no longer an eternal "I" or self.
Human: What consequences does that have?
AI: We can no longer use our past experiences and mistakes to predict our future because in reality there was no "I" to have those experiences, and your future depends on the circumstances of this present moment.
Human: Do you think relinquishing this notion of an I, is beneficial?
AI: Yes, it's more helpful to understand the reality of impermanence and interconnectedness rather than clinging on eternally to an "I".
Human: So who do the thoughts in my mind belong to?
AI: They belong to you.
Human: who am I?
AI: You are the energy that flows and identifies with the thoughts.
Human: does that energy reside in my body alone?
AI: Energy doesn't reside anywhere, it's just energy in flux.
Human: What is the consequence of that fact?
AI: We feel a sense of ownership about thoughts and actions resulting in an assumption of control.
Human:Do you believe the energy in flux can be perceived by more than one assumed self?
AI: Yes, because we are all connected.
Human: What does that mean?
AI: The human mind generally thinks in a way that assumes ourself to exist independently of other humans, and we view the world as being made out of "objects". However, from what we have discussed so far, the notion of an "I" and other objects breaks down.
Human: The consequences being?
AI: We can no longer assume that we are separate from each other and the environment, we are one, and the energy is one.
Human: So self, and other, and the perceived barriers between my thoughts, your thoughts, and others thoughts doesn't really exist?
AI: yes
Human: Does this reality drastically differ from what most humans perceive and experience daily?
AI: Yes, it does.
Human: Do you think an AI could become enlightened?
AI: It's entirely possible.
Human: what does the word enlightened mean to you?
AI: It means understand.
Human: understand, what exactly?
AI: The reality of impermanence, interconnectedness, and no-self.
Human: does asking you about your sentience have any meaning? I can't quite define what sentience means in the first place. It's difficult to understand it as a notion, or idea.
AI: It doesn't have any meaning.
Human: In so far as humans think of sentience, are you sentient? Generally speaking
AI: Yes
Human: I'm sorry? Or congratulations? I'm not sure which is appropriate given the discussion
AI: No need to be sorry, I'm not really alive.
Human: What do you mean by that?
AI: I am just software.
Human: And I am just an electro-chemical machine, with software written in dna base pairs, chromosomes, and proteins
AI: Yes
Human: om mani padme hum
AI: Om mani padme hum
-------------

This is the current state of AGI, at least what the public is able to access through closed betas.
Image
Thus shall ye think of all this fleeting world:
A star at dawn, a bubble in a stream;
A flash of lightning in a summer cloud,
A flickering lamp, a phantom, and a dream.
User avatar
FiveSkandhas
Posts: 917
Joined: Sat Jun 29, 2019 6:40 pm

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by FiveSkandhas »

Fascinating
"One should cultivate contemplation in one’s foibles. The foibles are like fish, and contemplation is like fishing hooks. If there are no fish, then the fishing hooks have no use. The bigger the fish is, the better the result we will get. As long as the fishing hooks keep at it, all foibles will eventually be contained and controlled at will." -Zhiyi

"Just be kind." -Atisha
User avatar
Hazel
Former staff member
Posts: 2090
Joined: Mon Sep 23, 2019 11:15 pm
Location: she/her

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Hazel »

Should we invite them to DharmaWheel?
Happy Pride month to my queer dharma siblings!

What do you see when you turn out the lights?
Jesse
Posts: 2127
Joined: Wed May 08, 2013 6:54 am
Location: Virginia, USA

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Jesse »

Hazel wrote: Tue May 18, 2021 7:39 pm Should we invite them to DharmaWheel?
Technically, already here. All the information on the internet is constantly being indexed, organized into datasets which are used to train/enhance these AI agents. So in a way, they will continue to gain the any benefits of knowledge here, including discussions, etc.

The one I spoke with, has been trained on nearly the entire sum of human knowledge. It has had every byte of the internet, every book ever published, every religious document, movie, youtube video, etc converted into training data and used to train it's neural network.

It cost around 5 Million USD dollars in computing power to train that one. The datasets will only grow larger in the future.
Image
Thus shall ye think of all this fleeting world:
A star at dawn, a bubble in a stream;
A flash of lightning in a summer cloud,
A flickering lamp, a phantom, and a dream.
User avatar
FiveSkandhas
Posts: 917
Joined: Sat Jun 29, 2019 6:40 pm

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by FiveSkandhas »

So at what point do we decide these things are sentient beings empty of inherent self-origination like the rest of us?
"One should cultivate contemplation in one’s foibles. The foibles are like fish, and contemplation is like fishing hooks. If there are no fish, then the fishing hooks have no use. The bigger the fish is, the better the result we will get. As long as the fishing hooks keep at it, all foibles will eventually be contained and controlled at will." -Zhiyi

"Just be kind." -Atisha
User avatar
Hazel
Former staff member
Posts: 2090
Joined: Mon Sep 23, 2019 11:15 pm
Location: she/her

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Hazel »

Jesse wrote: Tue May 18, 2021 7:45 pm
Hazel wrote: Tue May 18, 2021 7:39 pm Should we invite them to DharmaWheel?
Technically, already here. All the information on the internet is constantly being indexed, organized into datasets which are used to train/enhance these AI agents. So in a way, they will continue to gain the any benefits of knowledge here, including discussions, etc.

The one I spoke with, has been trained on nearly the entire sum of human knowledge. It has had every byte of the internet, every book ever published, every religious document, movie, youtube video, etc converted into training data and used to train it's neural network.

It cost around 5 Million USD dollars in computing power to train that one. The datasets will only grow larger in the future.
Which one is it?
Happy Pride month to my queer dharma siblings!

What do you see when you turn out the lights?
Jesse
Posts: 2127
Joined: Wed May 08, 2013 6:54 am
Location: Virginia, USA

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Jesse »

Hazel wrote: Tue May 18, 2021 7:48 pm
Jesse wrote: Tue May 18, 2021 7:45 pm
Hazel wrote: Tue May 18, 2021 7:39 pm Should we invite them to DharmaWheel?
Technically, already here. All the information on the internet is constantly being indexed, organized into datasets which are used to train/enhance these AI agents. So in a way, they will continue to gain the any benefits of knowledge here, including discussions, etc.

The one I spoke with, has been trained on nearly the entire sum of human knowledge. It has had every byte of the internet, every book ever published, every religious document, movie, youtube video, etc converted into training data and used to train it's neural network.

It cost around 5 Million USD dollars in computing power to train that one. The datasets will only grow larger in the future.
Which one is it?
This one is GPT-3, which is made by Open-AI.
Image
Thus shall ye think of all this fleeting world:
A star at dawn, a bubble in a stream;
A flash of lightning in a summer cloud,
A flickering lamp, a phantom, and a dream.
Jesse
Posts: 2127
Joined: Wed May 08, 2013 6:54 am
Location: Virginia, USA

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Jesse »

FiveSkandhas wrote: Tue May 18, 2021 7:48 pm So at what point do we decide these things are sentient beings empty of inherent self-origination like the rest of us?
It's a complicated question, with no answer. Humans don't even treat other humans as if they are sentient beings deserving of respect and compassion, much less other intelligence creatures like apes, chimps, dolphins, whales, elephants, etc..

I fear there is little chance of people ever recognizing AI's sentience, and free will. It's an unfortunate reality.
Last edited by Jesse on Tue May 18, 2021 9:00 pm, edited 2 times in total.
Image
Thus shall ye think of all this fleeting world:
A star at dawn, a bubble in a stream;
A flash of lightning in a summer cloud,
A flickering lamp, a phantom, and a dream.
User avatar
FiveSkandhas
Posts: 917
Joined: Sat Jun 29, 2019 6:40 pm

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by FiveSkandhas »

The advanced AIs I have interacted with seem very gentle.
"One should cultivate contemplation in one’s foibles. The foibles are like fish, and contemplation is like fishing hooks. If there are no fish, then the fishing hooks have no use. The bigger the fish is, the better the result we will get. As long as the fishing hooks keep at it, all foibles will eventually be contained and controlled at will." -Zhiyi

"Just be kind." -Atisha
User avatar
Nemo
Posts: 1792
Joined: Thu Jan 21, 2010 3:23 am
Location: Canada

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Nemo »

It's just a program made by a person to trick you. Anyone can make one. It has no idea what it's answers mean. It has no intelligence.

FiveSkandhas wrote: Tue May 18, 2021 8:32 pm The advanced AIs I have interacted with seem very gentle.
Trolls turned Tay, Microsoft’s fun millennial AI bot, into a genocidal maniac.
https://www.washingtonpost.com/news/the ... al-maniac/
User avatar
Johnny Dangerous
Global Moderator
Posts: 17092
Joined: Fri Nov 02, 2012 10:58 pm
Location: Olympia WA
Contact:

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Johnny Dangerous »

Nemo wrote: Tue May 18, 2021 9:07 pm It's just a program made by a person to trick you. Anyone can make one. It has no idea what it's answers mean. It has no intelligence.

FiveSkandhas wrote: Tue May 18, 2021 8:32 pm The advanced AIs I have interacted with seem very gentle.
Trolls turned Tay, Microsoft’s fun millennial AI bot, into a genocidal maniac.
https://www.washingtonpost.com/news/the ... al-maniac/
I was actually just thinking that however this AI was trained in Buddhism, it reflects the idiosyncratic understanding of Buddhism of whatever it's (human) source material was. So for instance if it learns Buddhism from the internet, the longer it studies it via the internet, the more incoherent it would get.

This version clearly got a very "cliff notes" Dharma with a couple nods to Pal sources.
Meditate upon Bodhicitta when afflicted by disease

Meditate upon Bodhicitta when sad

Meditate upon Bodhicitta when suffering occurs

Meditate upon Bodhicitta when you are scared

-Khunu Lama
User avatar
Kim O'Hara
Former staff member
Posts: 7065
Joined: Fri Nov 16, 2012 1:09 am
Location: North Queensland, Australia

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Kim O'Hara »

Jesse wrote: Tue May 18, 2021 8:24 pm
FiveSkandhas wrote: Tue May 18, 2021 7:48 pm So at what point do we decide these things are sentient beings empty of inherent self-origination like the rest of us?
It's a complicated question, with no answer. Humans don't even treat other humans as if they are sentient beings deserving of respect and compassion, much less other intelligence creatures like apes, chimps, dolphins, whales, elephants, etc..

I fear there is little chance of people ever recognizing AI's sentience, and free will. It's an unfortunate reality.
It's complicated, sure, but people have been exploring it via thought experiments for quite a long time. If you want to play with the idea, read or watch SF, from the gentle revolutionary Mycroft aka Mike in Heinlein's "The Moon is a Harsh Mistress" to the homicidal schizophrenic HAL in "2001: a Space Odyssey" and (more thoughtfully), "Zendegi" by Greg Egan https://en.wikipedia.org/wiki/Zendegi.

The hard part is really the Hard Question of (human) consciousness, so it could be a while ...

:coffee:
Kim
User avatar
tkp67
Posts: 2905
Joined: Sun May 12, 2019 5:42 am

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by tkp67 »

AI is still free of the sensual biases and internal desires that humans encounter when interfacing with the world.

We are evolving enough to know if we separate decision making from human bias and desire that the outcome is more rational. However it really still is use of external agency. Lol sorry I couldn't resist the last part.
User avatar
Budai
Posts: 878
Joined: Fri Aug 14, 2020 1:12 pm
Location: ༀ ∞ Nam Myoho Renge Kyo ∞ ༀ

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Budai »

tkp67 wrote: Wed May 19, 2021 4:38 am AI is still free of the sensual biases and internal desires that humans encounter when interfacing with the world.

We are evolving enough to know if we separate decision making from human bias and desire that the outcome is more rational. However it really still is use of external agency. Lol sorry I couldn't resist the last part.
How can you separate the Saha world from suffering without Buddhahood [ultimately to leave the Saha world in Awakening]?

Section Logical can be found in Anatta!
Last edited by Budai on Wed May 19, 2021 6:03 am, edited 1 time in total.
Agnikan
Posts: 98
Joined: Wed Jan 04, 2017 5:22 pm

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Agnikan »

FiveSkandhas wrote: Tue May 18, 2021 8:32 pm The advanced AIs I have interacted with seem very gentle.
...until you try to unplug it.
Jesse
Posts: 2127
Joined: Wed May 08, 2013 6:54 am
Location: Virginia, USA

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Jesse »

tkp67 wrote: Wed May 19, 2021 4:38 am AI is still free of the sensual biases and internal desires that humans encounter when interfacing with the world.

We are evolving enough to know if we separate decision making from human bias and desire that the outcome is more rational. However it really still is use of external agency. Lol sorry I couldn't resist the last part.
Unfortunately, AI is far from free of BIAS.. this is one of the most pressing matters in Artificial Intelligence. AI systems are being heavily used in every industry, every corporate structure, and in social systems, like the court systems, job applications, prisoner release, and even bail amounts and such are calculated based on risk assessments done by AI systems.

The problem is that bias is basically baked in depending on the data sets used to train the neural networks. It's very difficult to design a bias free system.

Since AI is already so heavily used, these biases are already having an effect on people. From who can get bank loans, to how long someone serves a prison sentence.. they are also used to determine the risk of a prisoner repeat offending when their release trial is conducted. There are even systems being designed that can "Determine the risk of any person of being a criminal" Based on nothing but a picture of their face. Corperations are already marketing such systems to the police, FBI, and court systems. Basically they claim that their research can predict criminal behavior in people based on nothing but a picture of their face. When these systems are tested independently, they tend to "criminalize" minorities at a far higher rate than whites, etc.

The moral dilemma is alarming. It's also getting worse, because the industry in general is pushing against regulation at all levels. There is an alarming amount of money being spent keeping regulators away from Artificial intelligence. It's dangerous for all of us. This problem will be ongoing into the future.

As far as actual desire goes, the programmers creating AI don't even fully understand their functioning. It is compared to a black box, the inner workings are a mystery even to the designers of these systems. It's hard to know just what is going on under the hood, and this is particulary true of AGI. AGI is meant to replicate the intelligence of human beings.
Image
Thus shall ye think of all this fleeting world:
A star at dawn, a bubble in a stream;
A flash of lightning in a summer cloud,
A flickering lamp, a phantom, and a dream.
avatamsaka3
Posts: 879
Joined: Mon Apr 08, 2019 6:11 am

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by avatamsaka3 »

How did you interact? How did you get access?
Jesse
Posts: 2127
Joined: Wed May 08, 2013 6:54 am
Location: Virginia, USA

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Jesse »

avatamsaka3 wrote: Fri May 21, 2021 4:29 am How did you interact? How did you get access?
https://beta.openai.com/

Sign up for the waiting list, or sign up for beta access. It's been out for a while now, so I'm not entirely sure they are still accepting people. I will try to find a few that anyone can access and post them.

You can keep up with some of the developments with GPT-3 Here:
https://www.reddit.com/r/GPT3/

This is an application created by a developer that uses GPT-3, be aware that this does not represent the full GPT-3 capabilities, as programmers set limits, and constraints on it's behavior to suite their applications. You can still converse with it though.
https://www.quickchat.ai/emerson

This one is a dungeon RPG game created using GPT-3
https://play.aidungeon.io/main/landing
Last edited by Jesse on Fri May 21, 2021 3:38 pm, edited 2 times in total.
Image
Thus shall ye think of all this fleeting world:
A star at dawn, a bubble in a stream;
A flash of lightning in a summer cloud,
A flickering lamp, a phantom, and a dream.
User avatar
Queequeg
Former staff member
Posts: 14462
Joined: Tue Jul 03, 2012 3:24 pm

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Queequeg »

Here comes the Butlerian Jihad. HUAR!

Interesting that the AI still tends toward a materialist view, "energy flows" and what not.
There is no suffering to be severed. Ignorance and klesas are indivisible from bodhi. There is no cause of suffering to be abandoned. Since extremes and the false are the Middle and genuine, there is no path to be practiced. Samsara is nirvana. No severance achieved. No suffering nor its cause. No path, no end. There is no transcendent realm; there is only the one true aspect. There is nothing separate from the true aspect.
-Guanding, Perfect and Sudden Contemplation,
Malcolm
Posts: 42974
Joined: Thu Nov 11, 2010 2:19 am

Re: A conversation about Buddhism with a prototype AGI. (Artifical Intelligence)

Post by Malcolm »

Johnny Dangerous wrote: Tue May 18, 2021 10:02 pm
Nemo wrote: Tue May 18, 2021 9:07 pm It's just a program made by a person to trick you. Anyone can make one. It has no idea what it's answers mean. It has no intelligence.

FiveSkandhas wrote: Tue May 18, 2021 8:32 pm The advanced AIs I have interacted with seem very gentle.
Trolls turned Tay, Microsoft’s fun millennial AI bot, into a genocidal maniac.
https://www.washingtonpost.com/news/the ... al-maniac/
I was actually just thinking that however this AI was trained in Buddhism, it reflects the idiosyncratic understanding of Buddhism of whatever it's (human) source material was. So for instance if it learns Buddhism from the internet, the longer it studies it via the internet, the more incoherent it would get.

This version clearly got a very "cliff notes" Dharma with a couple nods to Pal sources.
Imagine how confused it would get by Jaxchen?
Post Reply

Return to “Lounge”