[FRIAM] AI art

[Pieter Steenekamp]

I truly don't claim to know where AI is headed. However, I am convinced
that tomorrow's AI will be significantly smarter than today's. My
perspective on the future of AI isn't based on any special insight or
predictive power, but rather on my hopes for what could happen. I envision
AI developing to be far superior to humans in every aspect and fostering a
very positive relationship with us—akin to the relationship I have with my
children and grandchildren.

I'm quite old now, but for nearly half a century, I enjoyed playing tennis
with my son. Except for when he was very young, he has always been much
better than me—he's a skilled player, while I'm average. We created a
custom handicap system that allowed me to occasionally win, and I would
then make a big fuss about it, proudly telling the family that I had beaten
him. Everyone knew it wasn't a truly fair victory, but we all enjoyed it
immensely.

Lately, I've been teaching my 9-year-old grandchild to program in Python.
For now, I'm (probably) still better at it than him, but I truly enjoy
watching him develop and knowing he'll soon surpass me. We do this online,
as we live on different continents, and my daughter tells me how much he
enjoys our weekly Python sessions. My wife always reminds me of how much
joy it brings me.

I find true joy in seeing my offspring outperform me. My "wishful thinking"
prediction for the future relationship between humans and AI is that we
will find genuine happiness in our AI creations surpassing us in all areas.

On Mon, 24 Jun 2024 at 21:31, glen <gepropella at gmail.com> wrote:

> Exactly. The (relatively) successful AIs also have 2 components to their
> lineage: structural and tuning. The self-attending transformer architecture
> is analogous to either our anatomy or the ontogeny of the anatomy from DNA
> - structural lineage. I'd argue it's more like the former. The analogy to
> cultural descent is more difficult to iron out. But we can imagine
> something like LoRA, where some of the weights are more stable than others
> during any given training period ... like trying to unlearn idiom or speak
> with a different accent might be more difficult than learning new facts or
> "theorems". Some of the weights are subject to more frequent tuning than
> others. Etc.
>
> All we need to do is reverse engineer a DNA-like grammar from which to
> grow different architectures and we have something akin to biological
> evolution. Then develop a persistent MMO world allowing inter-player
> stigmergy and we have something akin to cultur[e|al evolution].
>
> On 6/24/24 11:55, Barry MacKichan wrote:
> > Our brains start off prewired to a significant if poorly understood
> degree. And then we learn from the full range of human experience in all
> its serendipitous contingency. We learn from the feel of an embrace and
> taste of ice cream, from battle wounds and wedding ceremonies and athletic
> defeats, from bee stings and dog licks, from watching sunsets and riding
> roller coasters and reading Keats aloud and listening to Mozart alone.
> Trying to train a computer about the meaning of love or grief is like
> trying to tell a stranger about rock and roll.
> >
> > “And not just from all these life experiences of our own but from the
> experiences of all our cultural ancestors. Our teachers—all those from whom
> we have learned, and those departed souls who taught our teachers—have
> shaped all those experiences into a structure of life, a system of values
> and ideals, a way in which we see and interpret the world.”
>
> --
> ꙮ Mɥǝu ǝlǝdɥɐuʇs ɟᴉƃɥʇ' ʇɥǝ ƃɹɐss snɟɟǝɹs˙ ꙮ
>
> -. --- - / ...- .- .-.. .. -.. / -- --- .-. ... . / -.-. --- -.. .
> FRIAM Applied Complexity Group listserv
> Fridays 9a-12p Friday St. Johns Cafe   /   Thursdays 9a-12p Zoom
> https://bit.ly/virtualfriam
> to (un)subscribe http://redfish.com/mailman/listinfo/friam_redfish.com
> FRIAM-COMIC http://friam-comic.blogspot.com/
> archives:  5/2017 thru present
> https://redfish.com/pipermail/friam_redfish.com/
>   1/2003 thru 6/2021  http://friam.383.s1.nabble.com/
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20240625/224acc55/attachment.html>