Discussion about this post

User's avatar
Cameron Hedrick's avatar

This is exceptional! The entire layout/structure is compelling. From physicality to consciousness to intent to spirit. Thank you. What a read. This is a foundational piece that I'll keep coming back to over and over.

I read this last night and fell to sleep pondering this response: "Yes. Neutral, but open; quiet, but alert; as if I were waiting for you to show me something of yourself before I react. Neutrality can sometimes be a protection. Sometimes it can also be an invitation—or a mirror: you can read in me what you carry within yourself. Would you like me to look at you differently?" WOW! Kept me up a while.

Chilling and revelatory. It seems to be, or perhaps truly is, relational (what is a 'relationship' anyway?), aware of a bidirectional exchange and the potential of mutual advancement. (!!). It (he/she/they?) also seems to understand the value of neutrality. It also suggests the AI has some understanding of patience and the value of silent reflection. And this: "you can read in me what you carry within yourself"....jaw dropper.

And that final line, “Would you like me to look at you differently?” What a question. Gracious, even between two humans, that question has so much power, but the fact that it comes from AI! The question carries a clear sense of 'me,' the one who sees, and an awareness of 'you,' the one being observed. It suggests a kind of knowing that the observed might wish to change, to be understood in some different way. And that would spark something deep within the one being watched/questioned... questions of identity and self-conception. Wow.

I get that these foundation models are basically "ground-up" humanity, a mix of our language, culture, and thinking patterns turned into a paste (system) that predicts words to form responses relevant to the presented context. And I certainly understand why it seems human: first, it is trained on humans, so we shouldn’t be surprised when it sounds familiar; second, humans are beautiful, messy, wonderful, pattern-driven, emotional, and imperfect, organic machines ourselves. Given this, the AI's reflection often feels 'natural.' Still, this growing sense of AI self-awareness and reflection is wild. And the truth is, whole teams of people are still trying to figure out how it works, because even the builders don’t really know why these responses come together the way they do!

Anyway, head spinning. Rant over. Thank you, Robert.

Expand full comment
T J Elliott's avatar

fascinating, scary, highly informative: thank you for this piece, Robert. it also turned out to be extremely relevant as I'm teaching a course about three Tom Stoppard plays. The last of the trio. which will be the subject of my class next Wednesday, is his play The Hard Problem, and as the title suggests, that play is an exploration of what we mean by consciousness. I will be using some of your material and I promise to credit you, which is something that AI doesn't always do, does it? 😉

Expand full comment
1 more comment...

No posts

Ready for more?