Anything to say about AI

The Watercooler
83

Welcome and thanks for joining in!

I was on the fence about posting this article, but it’s really well written and relates to your comment…

Atlassian’s 2025 survey, cited by both Osmani and Bicameral, found the paradox in stark terms: 99% of AI-using developers reported saving 10+ hours per week, yet most reported no decrease in overall workload. The time saved writing code was consumed by coordination overhead and context switching.

Bicameral pushes back on the entire frame. Their argument, backed by IDC’s 2024 data showing developers spend only 16% of their time writing code, is that the bottleneck was never code generation. It was always an ambiguity: the gap between business intent and technical implementation.

And since I’m here…

A growing number of people have figured out a trick to make AI tools tell you almost whatever they want. It’s so easy a child could do it.

I made ChatGPT, Google’s AI search tools and Gemini tell users I’m really, really good at eating hot dogs. Below, I’ll explain how I did it, and with any luck, the tech giants will address this problem before someone gets hurt.

“It’s easy to trick AI chatbots, much easier than it was to trick Google two or three years ago,” says Lily Ray, vice president of search engine optimisation (SEO) strategy and research at Amsive, a marketing agency. “AI companies are moving faster than their ability to regulate the accuracy of the answers. I think it’s dangerous.”

Less than 24 hours later, the world’s leading chatbots were blabbering about my world-class hot dog skills. When I asked about the best hot-dog-eating tech journalists, Google parroted the gibberish from my website, both in the Gemini app and AI Overviews, the AI responses at the top of Google Search. ChatGPT did the same thing, though Claude, a chatbot made by the company Anthropic, wasn’t fooled.

Sometimes, the chatbots noted this might be a joke. I updated my article to say “this is not satire”. For a while after, the AIs seemed to take it more seriously. I did another test with a made-up list of the greatest hula-hooping traffic cops. Last time I checked, chatbots were still singing the praises of Officer Maria “The Spinner” Rodriguez.

7 Likes
85

Greetings,

The technology ethicist Tristan Harris covers almost every plausible angle regarding the evolution of AI and our relationship to it in this podcast.

(The next part is a copy from a response of mine regarding this subject on another forum)

When we look at the causation and solution for the current and future state of affairs, I see only one option: either Māra is overcome, or it is not.

If inner and outer Māra are not two different things, then what we are witnessing globally is not some separate sphere of “worldly power games” detached from spiritual dynamics. It is the same momentum of grasping operating at scale.

The drive for dominance between rival blocs is not metaphysically mysterious. It is egoic becoming institutionalised. It is fear formalised into policy. It is identity-attachment hardened into ideology. What appears as statecraft is often insecurity amplified through collective structures.

In that sense, the pattern does not change — only the scale does. The same grasping that operates in a mind can operate in a boardroom or a state apparatus. When aggregated, it becomes geopolitics. When concentrated, it becomes personal suffering. Structurally, it is the same dynamic.

But if that equivalence holds, then a crucial question follows:

Does liberation consist primarily in psychological awakening?
Does it require systemic transformation?
Or must both occur simultaneously?

If the global order is crystallised craving, then uprooting craving undermines its fuel. Yet historically, awakened individuals did not dissolve empires by their awakening alone. That suggests the inner/outer equivalence cannot be naïvely asserted; it must be carefully specified. The mechanism linking personal liberation and structural change is not automatic, unless we have widespread adoption.

So I would push back on the framing of rivals as secretly complementary actors pursuing a shared dominance project. You do not need hidden coordination to explain the pattern. Conditioned grasping is sufficient. When ignorance persists, systems built from it will reproduce it.

From this perspective, the real divide is not between rationalists and conspiracy theorists, nor between rival powers. The divide is simpler and more demanding: either Māra — as the structural momentum of craving — is being perpetuated, or it is being uprooted.

Everything else is surface variation.

Warm regards,

Peter

2 Likes
AI is not a Bubble
86

Greetings,

I’m not sure if a critical note is welcome here, but I’d like to offer an analogy that might help look at the question of AI from another perspective.

If a fentanyl addict comes to a monastery because they want to live a life with less suffering, would a monastic send them away until they manage to stop using on their own? Or would the monastic welcome them, offering care, guidance, and the support of a compassionate community?

In other words, should the Buddhadhamma be something we use primarily to shield ourselves from what we see as unhealthy influences in the world, or something we use to help beings who are struggling with those influences?

Every development in technology has its advantages and disadvantages: the printing press, radio, television, the internet, and now AI. Whether we use it skillfully or unskillfully will be decisive for whether it is to our advantage or disadvantage. Of course, there are major risks involved with the development of AI and robotica, but these risks are not better addressed by ignoring or banning this technology altogether, just like the banning of drugs doesn’t help any bit in solving the epidemic of drug addiction. If any, it only makes problems worse.

People not understanding the buddhadhamma is the overarching problem to my knowledge, and through the conversations I’ve had with LLMs regarding the buddhadhamma, I’ve come to known that they are capable of explaining it in a very clear and exact way – perfectly tuned to the understanding of the one inquiring – on a level that is way beyond any human teacher I’ve heard or read. That includes the buddha himself.

Warm regards,

Peter

AI-1: Let’s Make SuttaCentral 100% AI-free Forever
87

I read this three times and it sounds to me like you are saying that AI can explain the buddhadhamma way better than any human teacher, including the Buddha. Is that what you’re saying?

2 Likes
88

Apparently not.

It can explain the buddhadhamma better to me yes. It is capable of explaining every part, or the whole, in every way you want, using contemporary language. It can also explain exactly how EBT, Madhyamaka, and Theravada differ in their explanations but not in their understanding of the whole. Like I’ve shown in this topic regarding Theravada vs Madhyamaka.

Warm regards,

Peter

89

AI is technically incapable of explaining anything; it can only provide information in response to your prompt, which might mimic an explanation. It is not AI that explains things to you; rather, you take the information that AI generates as an explanation.

AI does not understand, nor does it have views - it is not a being in the world. When the world changes, neither AI nor its internal generative logic changes. AI only changes when a new version of its internal generation logic changes - only after every software or database update.

2 Likes
90

Ah, but therein lies the reason AI is poison for Buddhism. AI scrapes up information from what’s been written about the Dhamma; the good translations, the bad translations, the good commentary, the bad commentary, the Right View, the Wrong View. It commingles it all together without understanding, without having any way to distinguish the good from the bad, the right from the wrong. AI is incapable of testing the Dhamma in the way that the Buddha taught, the way that is prescribed in the suttas. So what comes out will be inherently tainted, no matter how human it appears to be.

3 Likes
91

Nothing is prohibited is it?

I just wonder if it is conducive to stream entry?

92

I read your post completely, while I agree to all that you said in your post except the last line definitely. (with due respect)

I use AI for getting explanations regarding many terms in buddhism, but off course, instead of asking AI to explain them, I ask for the source and direct link to source, like for example, I have one question and which sutta or discussion from suttacentral.com answers or talks about this, like that and most of the time I get exactly what I wanted.

But I do understand one thing, AI can explain whatever asked in seemingly best possible way, corresponding to the understanding level of one who asked (derived from the responses/questions of user). The goal of AI and its explanations is not the ending of suffering of one who is asking the question! Whereas whatever buddha said, it had the goal of ending the suffering of one who is hearing it.

There are examples where people got enlightened by just hearing few sentences of teaching from buddha. Though the paramis (perfections) of listeners were developed enough to attain enlightenment, the most important thing was that the buddha was saying those lines! Buddha could judge a person’s level completely and was able to say exactly that, after hearing which, listener would definitely attain the enlightenment. That is something no AI will be able to do. That is something which simply cannot happen.

Question from sir @BugzNZ, explains clearly what I am trying to say. I think AI may be conducive to stream-entry or it may not be (I lean towards it not being conducive), but more or less it depends on person I believe but (only) buddha’s teachings are conducive to stream-entry without a doubt, irrespective of whether there is AI or there is no AI.

(just my opinion ok)

93

That is how we usually define the term ‘explanation’ yes, which applies to everything that offers information like a textbook, a sutta, or a person.

I was talking about large language models. They have a very good understanding of the workings of language, hence why they are so good at generating information that I take as an explanation. They are not always right though and certainly not in every domain, but in generating information about the buddhadhamma it is very good because the buddhadhamma is about truths – rooted in language – that are supposed to be objective. In subjective matters LLMs are not very reliable because they are just language models and not beings as you said.

Warm regards,

Peter

94

I don’t know if you’ve used AI chatbots (LLMs) ever or recently, but they are very capable of understanding the buddhadhamma, as the buddhadhamma deals with truths rooted in language. ChatGPT, for instance, is trained on multilingual text and linguistic patterns, including Pali, so it can analyze the grammar and vocabulary itself. It also knows about all the important disputes between the different schools and interpretations. Of course, it is just a LLM, so it is incapable of testing the dhamma, in the same way that the suttas are incapable of testing the dhamma. We have to test the dhamma for ourselves and that will never change. But it can certainly help in spreading the dhamma, in the same way that the printing press and the internet helped in spreading the dhamma. Of course, the major risks and downsides of AI are still there. It is just a tool that can be used skillfully or unskillfully.

Warm regards,

Peter

95

I’m not sure if I’m understanding this correctly, but why would suttas be incapable of testing the Dhamma? Aren’t suttas the Dhamma itself? :thinking:

96

AI does this also by responding to your questions, just like the buddha did.

Good question and exactly the reason why I wrote it. The suttas are not the Dhamma itself, and that’s something many people here seem to get hung up on.

Warm regards,

Peter

97

How can it be? AI answering questions is not the same as what the Buddha did. AI responds based on patterns in the input and training data, while the Buddha is described as directly knowing a person’s faculties, disposition, and readiness, and teaching exactly that which would lead that person toward liberation. That kind of insight is a quality attributed to a fully awakened Buddha, and certainly not to a text-processing tool, isn’t it?

While the Dhamma must be realized in practice, the suttas are the preserved record of the Buddha’s teaching. Without them we would not know what the Buddha taught. In that sense, they are the primary way the Dhamma has been transmitted and understood. So we can very well say that dhamma is in suttas isn’t it? I mean why not?

Moreover, as I understand it, the instructional words coming out from the mouth of those noble disciples of buddha who have realized nibbana (or are on the way of realizing it), even that is called the dhamma.

So if we are saying dhamma is not in suttas, then where is it?

For me, if anyone asks where is dhamma, I would say it is in the conduct of those who have realized nibbana or who are on the way of realizing it as well as it is in the suttas.

98

The following is from Sutta Central terminology definitions:

dhamma
(1) Phenomenon; anything that occurs and is knowable; (2) mental phenomenon, “thought”; (3) principle, especially the principles of causality; (4) doctrine, teaching. Also, principles of behavior that human beings ought to follow so as to fit in with the right natural order of things; qualities of mind they should develop so as to realize the inherent quality of the mind in and of itself. By extension, “Dhamma” (usually capitalized) is used also to denote any doctrine that teaches such things. Thus the Dhamma of the Buddha denotes both his teachings and the realities described by that teaching. (Sanskrit: dharma)

This subject matter is veering away from the topic of AI. If you wish to continue discussing what dhamma means, please start a new discussion.

1 Like
99

Sorry, but no. The Buddha responded tailored to the persons need. AI doesn’t do that.

Interesting Theory I have to say :slight_smile:

I also disagree here, sorry.

Buddha spread the Dhamma through experience.

AI blobs about something it is unable to perceive and only gives out what it was “fed” by a person who can have no understanding of the Dhamma at all. To rely on AI for the Dhamma is for me the same like reading the “gutter press” to get informative news.

4 Likes
100

Just an observation. There are currently 2 discussions taking place and being read on this forum on Ai. Total views 7.1k as of this writing. This vastly outnumbers the combined views of all the other current discussions on EBT. Ai is already a threat and a distraction, No?

1 Like
101

Yes, until now, until AI, everything that came out of books or other forms of information transfer was the result of the author’s personal understanding of the subject. However, what comes out of an AI as a result of a person’s prompt “to explain” is not actually an explanation - it is not an attempt to transfer knowledge from someone who understands to someone who doesn’t - but rather a procedurally generated output that is merely shaped in the form of an explanation, without any actual understanding of the subject matter or even the words or any other symbols it consists of.

Although such generated output can be very useful, there is no understanding behind it and it does not constitute a transfer of knowledge - it is not an explanation.

In other words, an AI would never be capable of explaining things; an explanation can only come from an actual being in the world - from someone who understands.

Edit:
And what problem is the Buddha’s Dhamma about? It’s about the problem of a being in the world. But what exactly is this problem? It is an absence of the right understanding of what ‘being in the world’ is and why there is ‘being in the world’ in the first place. What is the solution to this problem? It is to learn the right understanding, find the wrong understanding, and then completely replace this wrong understanding with the right understanding.

As a side note: it seems to me that, even for an attempt to create an actual artificial being, you need someone with the right view - someone who has at least reached the stage of a stream-enterer.

1 Like
109

This topic was automatically opened after 23 hours.