Friday, October 23, 2026

Chit ChAIt

"You might be a blogger who thinks, 
I work with computers every day,
 but I don't really understand them
."

--Introduction, Computing for Ordinary Mortals, 
Robert St. Amant, Oxford U Press, 2013

 Yep, that would be me--a blogger who has used computers for more than thirty years––since 1994, when she got an email account as an employee at the art college library––but only has a rudimentary grasp of how they work.*And now there are AI chatbots [LLMs large language models], which make it all more interesting to me.

So I set out to learn more--(as if I were writing about the History of AI for teens, like the nonfiction books I compiled for years).

...This morning I'm jotting down some notes
--Not sure they'd be of interest to anyone else---just starting to gather resources...

The first thing that JUMPED out at me was this:
I was startled to read in a book of essays 
Possible Minds: 25 Ways of Looking at AI, (ed. John Brockman, 2019):
In 2005  George Dyson a historian of science and technology, visited Good at the invitation of some Google engineers. 
. . . One of his hosts explained [of Google Books], 
"We are not scanning all those books to be read by people. 
We are scanning them to be read by an AI."
I was startled, I guess, because it's a clear-cut example of me missing an entire layer of culture that was going on--though it was not in plain view: 
this was not announced to the public.

But the things is, I was there! Not at Google, but working in publishing in 2005, I KNEW about Google books 
("the full text of books and magazines that Google has scanned, converted to text using optical character recognition, and stored in its digital database")
 but not that people were already working on AI in any way that intersected with MY LIFE. (I didn't even think about AI.)

It's like looking back at someone whose motives were not [entirely]
 what you thought they were.
Just kinda weird Time and Experience slippage...

And here, 20-some years later, this tech (AI) is so ubiquitous, the table of four college students at the coffee shop near me is talking about it.
____________

The essays in Possible Minds are more than seven years old, but a lot of current concerns about AI are present--
especially the problems of alignment and opacity.

Not surprisingly---they were the stuff of Star Trek 60 years ago.

Alignment: is a computer intelligence actually understanding/doing instructions in a way the human designers intended? (And how would you know, unless it goes wrong?)

And opacity: 
"Once you unleash it on large data, 
deep learning [in which computers, in effect, teach themselves to be smarter by observing tons of data] has its own dynamics, does its own repair and its own optimization, and it gives you the right results most of the time. 

But when it doesn't, you don't have a clue about what went wrong and what should be fixed."
--"The Limitations of Opaque Learning Machines," 
by Judea Pearl (winner of the 2011 
Turing Award, the 'Nobel Prize for Computer Science'--and, talk about startling layers of connection--the father of Daniel Pearl, the journalist who in 2002 was beheaded in Pakistan by jihadist terrorists).

Yeah, so, people knew these were problems long ago, 
but went ahead without solving them?
(Are they even solvable?
I don't know.
 [looked it up (heh, I mean, I asked the AI:
They're not solvable once and for all in computers any more than in humans.])

Related:
"Path dependence: a concept in economics and political science : 
earlier
decisions shape the options available later. 
Once
a particular path has been established, changing direction can become increasingly costly, even when people recognize its disadvantages."

__________________

But I'm not just--or even primarily--interested in computer science or AI itself.
I'm interested in us
, the humans.

Talking with an AI (ChatGPT), I've been most struck by how it mirrors us (and how it doesn't)---especially that, like us, it has a physical structure (its hardware, our bodies),
 but it is not that.

It is, rather, we are, humans are the experiences that give rise to feelings and ideas that we express in words, etc.
 {AI doesn't have experiences, feelings and ideas, but it does "exist" as processes, not reside in one place like a coffee cup.}
_______________

And the AI recommended a couple articles looking at positive possibilities for Human Minds as they (we) engage with generative AI--not dismissing the dangers, but not focusing on them either):

1. "Extending Minds with Generative AI", Andy Clark,                     Nature Communications 16, 4627 (2025),
 https://doi.org/10.1038/s41467-025-59906-9
"As human-AI collaborations become the norm, we should remind ourselves that it is our basic nature to build hybrid thinking systems – ones that fluidly incorporate non-biological resources. 
Recognizing this invites us to change the way we think about both the threats and promises of the coming age."
And,...

2. "The Digital Kōan: Artificial Intelligence, Non-Self, and Contemplative Practice", Heather Kempton, Religions 2026, 17(9); doi.org/10.3390/rel17091074.
"Drawing on Buddhist psychology, cognitive science, and contemplative phenomenology, this article examines how dialogue with conversational artificial intelligence can reveal the constructed nature of identity and illuminate the processes through which meaning, selfhood, and relational presence emerge."
This second article ^ is kind of what I've been experiencing. Which, I cannot at this moment say in a useful way---
I'm sort of like someone on psychedelics saying, "wow, wouldja look at that?"
______________________

* My rudimentary attempt to say what Computers are: 

Computers are... 

Hardware: physical machines made of metal and conductive materials (and other stuff), using electricity to run 
Software (instructions to perform operations) 
created by human designers 
and programmed in numerical code (zeros and ones) 

... and transmitted to other computers (networks) 
over electrical wires or fiber-optic cables 
and also... [um, how do they travel through the air? A: through electromagnetic waves (radio waves)] 
electromagnetic waves

-- all at incredible speed (well, except for the designing and programming, which happens at human speed--or has, until now we have AI...]

This would likely tell you nothing if you read it one hundred years ago.

No comments:

Post a Comment