Getting Real with AI

3 months ago 3
The incorporeal non-place where we also live. By Hugh McLeod, 2004.

When I read that some conversations with ChatGPT had appeared in Google searches, I did a search for “Doc Searls” ChatGPT and got a long and not-bad but not entirely accurate AI summary below which normal-ish search results appeared. When I went back later to do the same search, the results were different. I tried the exercise again in another browser and got entirely different results. I also found no trace of personal chats with ChatGPT surfacing on Google. But with returns diminishing that fast, why bother?

What I did come to realize, quickly, is that there is no “on” anymore with Google. And there may never be with AI as we know it, and as it seems to be playing out.

There is no “on” in “online.” No “in.”

We use adpositions, which include prepositions, to make sense of the natural world. But they don’t truly apply (though we try) to the digital world. Because the digital world isn’t real in the natural or physical sense. See, adpositions, and prepositions especially, are made for our embodied selves in the natural world. Under, around, through, beside, within, beneath, above, into, near, toward, with, outside, amid, beyond, are all sensibe in the here where we eat and breathe. None of them are of the digital world. But that world is no less real for it. Cyberspace is beyond ironic. It is oxymoronic, self-contradictory. There is no space involved. When persons in Sydney, Lucerne, New York, and Tokyo are together talking and seeing each other on Zoom, they are still nowhere, because there is no where in the physical sense. We are adjacent across the non-space within a giant zero.

But we do live there. Together. And apart.

This all came to mind this morning when I read two pieces:

The first speaks to living disembodied lives along with our embodied ones.

The second speaks to the mania for Big AI spend:

It’s also worth breaking down where the money would be spent. Morgan Stanley estimates that $1.3tn of data centre capex will pay for land, buildings and fit-out expenses. The remaining $1.6tn is to buy GPUs from Nvidia and others. Smarter people than us can work out how to securitise an asset that loses 30 per cent of its value every year, and good luck to them.

Where the trillions won’t be spent is on power infrastructure. Morgan Stanley estimates that more than half of the new data centres will be in the US, where there’s no obvious way yet to switch them on.

I now think that money will be far better spent on personal AI.

That’s AI for you and me, to get better control of our lives in the natural world where we pay bills, go to school, talk to friends, buy goods, get sick and well, entertain ourselves and others, and live lives thick with data over which we have limited control. (Do you have any record of all your subscriptions, of your health and financial doings and holdings, of what you’ve watched on TV, of where you’ve been, and with whom? Wouldn’t it be nice to have and make sense of all that stuff? I mean by yourself and for yourself, and not off in the cloud of some giant who can do fuck-all with it?)

It’s as if we are back in 1975, but instead of starting to work on the personal computer, all the money spent on computing goes into making IBM and the BUNCH more gigantic than anything else ever, with spendings that dwarf what might be spent on simple necessities, such as the electric grid and roads without holes. Back then we had the good fortune of Jobs, Wozniac, Osborne, and other early mammals working on personal computing underneath the feet of digital dinosaurs. Do we have the same today? Name them. I’m curious.

No, I’m not talking about people working on better ways to buy stuff, or to navigate the digital world with the help of smart agents. I’m talking about people working on stuff you and I use to get control of our everyday lives and the data we need to manage—without the help of giants.

Like we started doing with PCs fifty years ago.

Read Entire Article