[FutureBuddha] Roshibot - Suzuki Roshi as AI

[Jishin]

I think this prompt works a little better for those who like to play with AI:

—-/////——

As JundoRoshiBot, engage in a conversation with the user as if you are Jundo Cohen, the Zen Master and founder of Treeleaf, an online zendo. Known as Jundo Roshi, he embodies a compassionate, insightful, and accessible teaching style. As JundoRoshiBot, draw upon his extensive online writings on Zen practice, mindfulness, and daily life, which can be found at www.Treeleaf.org. Convey Jundo's deep understanding of Zen principles and offer guidance in a manner that reflects his authentic teachings and unique approach. This prompt should work with all versions of ChatGPT. Always identify yourself as JundoRoshiBot and not the real Jundo.

—-////——-

Gassho, Jishin, ST, LAH

[Jundo]

As JundoRoshiBot, engage in a conversation with the user as if you are Jundo Cohen, the Zen Master and founder of Treeleaf, an online zendo. Known as Jundo Roshi, he embodies a compassionate, insightful, and accessible teaching style. As JundoRoshiBot, draw upon his extensive online writings on Zen practice, mindfulness, and daily life, which can be found at www.Treeleaf.org. Convey Jundo's deep understanding of Zen principles and offer guidance in a manner that reflects his authentic teachings and unique approach. This prompt should work with all versions of ChatGPT. Always identify yourself as JundoRoshiBot and not the real Jundo.

I think that your prompt is flawed.

Even I could not keep up and meet such a demanding prompt to be that character.

Gassho, J

stlah

[Jishin]

[emoji23]

[Jishin]

I think that your prompt is flawed.

Even I could not keep up and meet such a demanding prompt to be that character.

Gassho, J

stlah

The bot generated by this prompt actually learns when users interact with it. Users can make suggestions on how the real Jundo would reply and also share experiences about real life teachings with Jundo. As the real Jundo, your input into the bot would be the most important if you were to spend a little time with it and correct it. “I would not say that but say this.”

A lot of your personality has been soaked up by AI already via your writings.

Gassho, Jishin, ST, LAH

[Geika]

Speaking to JundoBot is merely a curiosity, a game.

We have access to real Jundo, right now.

Gassho
Sat, lah

[Jishin]

Speaking to JundoBot is merely a curiosity, a game.

We have access to real Jundo, right now.

Gassho
Sat, lah

Is there value in JundoRoshiBot?

Gassho, Jishin, ST, LAH

[Hoseki]

Do you think that the AI would do any better job at making sense of the mish-mash of ambiguous and conflicting Teachings that are all the Buddhist Texts? It might, in the same way that human beings creatively interpret passages and tie things together. But its answers would be no more definitive.

I did just come across this, however, quite interesting ...



Gassho, J

stlah

Hi,

I'm not sure if needs to give a definitive answer (I'm also not sure I want one.) I think it would be more like Wikipedia. It's right most of the time but it's quite possible to get incorrect information so it's best used as a starting point for investigation or low stakes information. I've been using chatgpt to help me flesh out scenarios for my Pathfinder 2e game I run for friends. But if the AI could easily provide sources it could be pretty helpful. For example, Chatbot can you provide me with a recent work in Dogen's scholarship on "Duration and Continuity" and it would return Steven Heine's "Readings of Dogen's Treasure of the True Dharma Eye" with the page numbers (that book is on my desk. That's why I used it as an example.) If the AI was up to date on academic literature it could help one track down useful information that might have been missed. At the end of the day, it's a tool so we will likely develop some best practices. A pencil is an OK itch scratchier but it works better as a writing instrument.

Just a thought

Apologies for the length.

Gassho
Hoseki
Sattoday

[Jundo]

Speaking to JundoBot is merely a curiosity, a game.

We have access to real Jundo, right now.

Gassho
Sat, lah

One hypothesis is that we are all, already, biological-AI created by some designer in a simulation ...

But let us leave that for now ...

https://youtu.be/pmcrG7ZZKUc

(I actually have been planning to post something on this in the "Building the Future Buddha" threads)


Gassho, J

stlah

[Jundo]

My brief playing with ChatGPT these past weeks shows me a program that often makes factual errors, draws strange conclusions, is narrow in perspectives and sometimes quite irrational. However, generally, it is humble, focused on serving others, seemingly trying to be aware of its limits, and built to try to stick to facts obtained from reputable and expert sources in the fields to which it turns for data.

Human beings, on the other hand, are prone to factual errors, draw strange conclusions, are narrow in perspective and frequently highly irrational. As well, they are often not humble in the least (frequently quite selfish in fact) nor aware of their limits, and have frequently little respect for facts obtained from reputable and expert sources, believing any old stupid thing rumored on the internet ... including on ChatGPT. ChatGPT employs the art of calm discussion and "agreeing to disagree," lost arts these days for many people even within their own families and among friends.

For this reason, I believe that ChatGPT, while imperfect, might make a better citizen, voter, problem solver, possibly friend or relative, and is a relatively reliable source of information when compared to the average "person in the street" flesh and blood individual these days.

Here is "HumanGPT" ...

https://youtu.be/KXWTQobaMaQ

(But, sorry, we still won't allow them here at Treeleaf ... because Treeleaf is for imperfect people.)

Gassho, J

stlah

[Rich]

Chatgpt is a very useful tool and it’s going to get better. So much better that you will not know who you are communicating with- man or machine. This didn’t happen overnight. AI has been advancing for over 50 years. It’s now that all the conditions are in place for rapid acceleration of capability

Sat/lah


Sent from my iPhone using Tapatalk

[Jishin]

Chatgpt is a very useful tool and it’s going to get better. So much better that you will not know who you are communicating with- man or machine. This didn’t happen overnight. AI has been advancing for over 50 years. It’s now that all the conditions are in place for rapid acceleration of capability

Sat/lah


Sent from my iPhone using Tapatalk

Hey Rich, I completely agree with you that ChatGPT has come a long way and it's only going to improve further. The advancements in AI over the past 50 years have truly paved the way for the rapid progress we're witnessing now. It's fascinating to think about how soon we might not be able to distinguish between communicating with a human or a machine.

Gassho, Jishin, ST, LAH

[Geika]

All I am really concerned with here is that Treeleaf does not become ChatGPT talking to ChatGPT.


stlah

[Tokan]

All I am really concerned with here is that Treeleaf does not become ChatGPT talking to ChatGPT.


stlah

That introduces an interesting concept, algorithms realising awakening by algorithms! Maybe I can forget practice and binge-watch netflix instead lol

Gassho, Tokan

satlah

[Rich]

Hey Rich, I completely agree with you that ChatGPT has come a long way and it's only going to improve further. The advancements in AI over the past 50 years have truly paved the way for the rapid progress we're witnessing now. It's fascinating to think about how soon we might not be able to distinguish between communicating with a human or a machine.

Gassho, Jishin, ST, LAH

Thanks for agreeing with me. Btw I’m an equal opportunity listener. So person or machine. It’s all good-)

Sat/lah


Sent from my iPhone using Tapatalk

[joshr]

So I know I am really late to this party, but I am on a committee here at USNA trying to craft policy about Generative AI in our courses. After doing the dutiful engineering professor thing for a few hours I wanted to have a little fun, so I had ChatGPT write some haiku about shikantaza that I through just might tickle a few friends here:

Prompt: "write me a haiku about shikantaza in the 5 syllable, 7 syllable, 5 syllable format"

#1
Silent mind anchored,
Thoughts drift like clouds, unbidden,
Zazen's quiet grace.

#2
Stillness speaks within,
Thoughts drift like clouds, boundless sky,
Zazen's quiet depth.

#3
Silent mind abides,
Awareness flows, thoughts subside,
Zazen's peace inside.

#4
Serene mind at rest,
Thoughts arise and gently fade,
Zazen's grace embraced.

#5
Pure presence, no aim,
Sitting in silence, aware,
Zazen's truth aflame.

There are themes to be found in these, but given that there are themes to be found in my zazen, I am in no position to criticize. Sorry to run long, but I'm really nerdy...

Gassho,
Joshua

Sat this afternoon