"The non-seeker is the last disguise of the seeker." What are we seeking? Are there no finders? Douglas Harding, I believe, would say to point your finger at your "face." What is the finger point at? That's it! You've found it. You are headless. You are the singular first person and have no head, not like the multitude of third persons, all with heads. I see Douglas as talking about the nature of perception and how to get it right. I'd like to see Douglas in one of these conversations. Not face to face with poeple who might have other viewpoints. There is no such thing as face to face. No one has ever been face to face with anyone. It's always face to no-face. Claude knows all about this no-face. I'd like to see the dialogue he creates.
I'm happy you've moved on from AI-generated conversations between my suggested list of characters to your own list. As with the previous ones, I feel that AI gets some things right, but often misses the mark. For example, it has K saying things he'd never have said. My sense is that AI gets the surface but not the depth, especially with someone like K, who pointed beyond words and concepts. The AI seems to get phrases and themes, at least superficially, but simply can't go where K is pointing. So it stays on the level of discussing ideas.
The artwork is disappointing. The only one of the 3 who resembles the real person is K. I'd never have guessed who the other 2 were supposed to be.
Finally, I like Sharon Hanna's suggestion to include the donkeys.
Anyway, I think I've had my fill of imaginary AI-generated dialogs between real people. I definitely prefer the real people. (And if anyone responds by saying there are no real people, they can consider themselves whacked over the head by an imaginary AI-generated Zen Master). 😎
The policy of OneAI (GPT's creator) does not allow accurate depictions of real people. I've found ways of working around that to some extent, but you are right. These are not among the best I have been able to generate.
As for the content, you seem to miss the miracle here. This entire conversation was written from a simple prompt telling the AI to write an imaginary conversation among Watts, Krishnamurti, and Saltzman. The AI then went to the internet to learn about those figures and constructed the entire thing with no input from me. All of it: Ojai, Jiddu's habit of interrupting, Watt's self-satisfied humor, my irony, etc. The AI can't know those people the way you can know them, but this does capture quite accurately some of their points of view. I find it amazing.
I understand that AI created all of this simply by looking over material on the internet, and it certainly did capture something about the ideas and forms of expression of these folks, and it was able to create a seemingly real dialog between them in which they discuss some of these ideas. Indeed, that is quite miraculous. And playing with this can certainly be entertaining and fun.
But in all the dialogs like this that you've prompted AI to do in recent days, including this one, the AI gets it wrong in many instances, putting words into the avatar's mouths that the real people would never have said, often getting them quite wrong, and it misses entirely the real depth to which they were often pointing. As I said, AI can never go where K was pointing.
I suggested the people I did for the previous dialogs you had the AI do because they represented the different (sometimes seemingly contradictory) nondual/spiritual threads or perspectives that have most deeply touched me and become part of my own outlook and expression, and I thought that having them converse might be revealing and even in some way helpful. And now it appears you've chosen two people that I know both touched you deeply (Watts and K) to have a dialog with each other and with you.
And doing this is fun and interesting to some degree, as long as we remember that these are imaginary conversations between AI-generated avatars and not dialogs between the real people they claim to represent. For me, the intrigue was short-lived, and the dangers and limitations were quickly apparent. But I get that you find it all very amazing, and that it seems to interest you in a way it doesn't (at least so far) interest me. But it was fun, for a while. 🙏❤️
As you know, Joan, I have been thinking deeply about AI and what its arrival means for humanity. Understanding Claude investigates what AI is and isn't. Tomorrow, I will publish a short essay here that goes to the essence of my findings to date and contains a warning for all us.
As a longtime Krishnamurti gal I thoroughly enjoyed this! Excellent! Thanks Robert!
I would like the donkeys to be characters, please. It could add some “devil may care”-ness. ….. all three. They could bray, etc.
🫏🫏🫏…..
"The non-seeker is the last disguise of the seeker." What are we seeking? Are there no finders? Douglas Harding, I believe, would say to point your finger at your "face." What is the finger point at? That's it! You've found it. You are headless. You are the singular first person and have no head, not like the multitude of third persons, all with heads. I see Douglas as talking about the nature of perception and how to get it right. I'd like to see Douglas in one of these conversations. Not face to face with poeple who might have other viewpoints. There is no such thing as face to face. No one has ever been face to face with anyone. It's always face to no-face. Claude knows all about this no-face. I'd like to see the dialogue he creates.
A fun experiment. An enjoyable read.
It's creative, it’s fun and funny. It could be a great theatre play.
Thank you, Robert.
I'm happy you've moved on from AI-generated conversations between my suggested list of characters to your own list. As with the previous ones, I feel that AI gets some things right, but often misses the mark. For example, it has K saying things he'd never have said. My sense is that AI gets the surface but not the depth, especially with someone like K, who pointed beyond words and concepts. The AI seems to get phrases and themes, at least superficially, but simply can't go where K is pointing. So it stays on the level of discussing ideas.
The artwork is disappointing. The only one of the 3 who resembles the real person is K. I'd never have guessed who the other 2 were supposed to be.
Finally, I like Sharon Hanna's suggestion to include the donkeys.
Anyway, I think I've had my fill of imaginary AI-generated dialogs between real people. I definitely prefer the real people. (And if anyone responds by saying there are no real people, they can consider themselves whacked over the head by an imaginary AI-generated Zen Master). 😎
Hi, Joan.
The policy of OneAI (GPT's creator) does not allow accurate depictions of real people. I've found ways of working around that to some extent, but you are right. These are not among the best I have been able to generate.
As for the content, you seem to miss the miracle here. This entire conversation was written from a simple prompt telling the AI to write an imaginary conversation among Watts, Krishnamurti, and Saltzman. The AI then went to the internet to learn about those figures and constructed the entire thing with no input from me. All of it: Ojai, Jiddu's habit of interrupting, Watt's self-satisfied humor, my irony, etc. The AI can't know those people the way you can know them, but this does capture quite accurately some of their points of view. I find it amazing.
I understand that AI created all of this simply by looking over material on the internet, and it certainly did capture something about the ideas and forms of expression of these folks, and it was able to create a seemingly real dialog between them in which they discuss some of these ideas. Indeed, that is quite miraculous. And playing with this can certainly be entertaining and fun.
But in all the dialogs like this that you've prompted AI to do in recent days, including this one, the AI gets it wrong in many instances, putting words into the avatar's mouths that the real people would never have said, often getting them quite wrong, and it misses entirely the real depth to which they were often pointing. As I said, AI can never go where K was pointing.
I suggested the people I did for the previous dialogs you had the AI do because they represented the different (sometimes seemingly contradictory) nondual/spiritual threads or perspectives that have most deeply touched me and become part of my own outlook and expression, and I thought that having them converse might be revealing and even in some way helpful. And now it appears you've chosen two people that I know both touched you deeply (Watts and K) to have a dialog with each other and with you.
And doing this is fun and interesting to some degree, as long as we remember that these are imaginary conversations between AI-generated avatars and not dialogs between the real people they claim to represent. For me, the intrigue was short-lived, and the dangers and limitations were quickly apparent. But I get that you find it all very amazing, and that it seems to interest you in a way it doesn't (at least so far) interest me. But it was fun, for a while. 🙏❤️
As you know, Joan, I have been thinking deeply about AI and what its arrival means for humanity. Understanding Claude investigates what AI is and isn't. Tomorrow, I will publish a short essay here that goes to the essence of my findings to date and contains a warning for all us.
"Now this is my kind of temple. Three minds talking past each other with great affection. Each convinced he’s not the one who’s confused."
"Written by ChatGPT 4.0." Really? This wasn't edited for style, clarity, or effect?
No edits. I gave the AI the list of characters and the suggested setting and it took it from there.
Impressive. My own speech is clearly generated by a naturally evolved LLM, an auto-complete system in a meat sack, an LLM on steroids. Literally.
More evidence, I guess, that the prattling idiot in my head isn’t “me.”
Thanks for sharing the article!
Yes. Daniel. That is a subtle takeaway from my work with LLMs. Understanding Claude goes into that theme in depth from various angles.