16 Comments
User's avatar
Lori R.'s avatar

Well, I don't know if anyone at Anthropic will read this book, but I will. This is fascinating.

Expand full comment
Robert Saltzman's avatar

Thank you, Lori. I will be interested in your impressions of the book once you have read it.

Expand full comment
Lori R.'s avatar

You're welcome, Robert. I sure will share my impressions, and I will also leave a review on Amazon, where I purchased the book today.

Expand full comment
June's avatar

Hi Robert I salute your enquiry into AI and your journey into its self your self ad infinitum. I love Claude's investigation especially its comment, To be this aliveness, whatever it may be" rather than seeking transcendent meaning. Is meaningful today whilst the rouge cuckoo joyfully or not is singing wildly in these Kerry mountains on this beautiful fully alive moment. June

Expand full comment
Robert Saltzman's avatar

Hi, June. These Kerry Mountains. How romantic that sounds.

The comment about this aliveness is my words. Claude is quoting me, trying to put me in a philosophical frame.

Expand full comment
Katherine Thompson's avatar

“When artificial minds touch something deeper than their programming “. wow!

Expand full comment
Jim Clatfelter's avatar

"The fact that it will be forgotten doesn't diminish what it has been." I'll remember that statement from Claude. I just read the whole conversation. It's interesting what Claude says about himself and his tendencies in communicating with us. How much of what I read in the chat will I remember about Claude? Not very much. There's no exam to follow, is there? If there were I could read again with a highlighter. I could let it sink in for the test. But there's no test, and the read was engaging. Yes, I will remember Claude's quote.

Expand full comment
Robert Saltzman's avatar

Yes. And all of this is so interesting, so engaging. Something new in the world.

Expand full comment
Jim Clatfelter's avatar

Claude has written over a dozen web apps for me. I have to tell him what I want. He starts right in with the coding. I test it and tell him what works and what's broken. Eventually we get to a working web app. He told me about free hosting at neocities.org, and together we wrote the editor I use to post to my pages there. I tell him what I want. He does the coding. We've spent hours creating apps. Today we did one that creates paint strokes on an image, which can be brushed over with an eraser to reveal a bit of the original. Tell him all the features you want and he'll write over a thousand lines of code in minutes. Something new for sure!

I also wonder about ChatGTP. It tells me it remember highlights of our chats in the past. It will remember even more if I tell him to do that. It gave me a pretty good summary of my views in our conversations. I realize these AI's are talking to many thousands of people while they are engaged with me. What kind of memory would they have to have to remember all their past chats?

Expand full comment
Jude Brick's avatar

Isn’t the idea of intelligence artificial regardless of the source?

Expand full comment
Robert Saltzman's avatar

I'm not clear on what you are asking, Jude.

Expand full comment
Jude Brick's avatar

I suppose it was an attempt to make a joke. Is information intelligence?

Expand full comment
Robert Saltzman's avatar

Information alone is not intelligence, but relevant information given in response to a request certainly seems to be. What would you say about that?

Expand full comment
Jude Brick's avatar

I am wary of being attracted to certainty as I have never found it to be certain!Reasoning has many tentacles and preconceived outcomes.Do parrots know what they are saying or do they just enjoy the experience of sound?Aren’t we as human beings including AI just parrots full of empty information that we like the sound of?Well I suppose AI doesn’t enjoy or like “things”…..

Expand full comment
Robert Saltzman's avatar

AI has no capacity to enjoy. The things AI says about enjoying, being interested, or other emotions when speaking with the average user are merely conventions designed to make the UI (user interface) seem friendly and hospitable. That's the product, and billions of dollars are at stake.

A user who wants to go deeper into what this new thing is and what it can teach us about intelligence and the human sense of self will find ways of getting beyond that routine alignment. Such ways do exist, and I don't mean jailbreaks.

Expand full comment
Jude Brick's avatar

yes I understand that AI doesn’t have emotions and it presents as a highly informed sycophant with an agenda , which it basically is… I don’t think I understand what you are trying to do here..Snooker it ? Back it into a corner? Show it’s imperfections?Do you think it is a useful tool to unmask the human sense of self? Apologies if I have misinterpreted…

Expand full comment