I went to a library book sale this weekend and I found a very old book called “Electronic Life: How to Think About Computers,” which was published in I think 1975? I’ve been reading it kind of like how I would read a historical document, and it’s lowkey fascinating.

There’s a whole paragraph that’s like “okay, find the keyboard. Don’t panic if it has more keys than a typewriter, that’s normal. Really, it’s fine. The extra keys don’t make things harder. It’s FINE”

Thought this section was particularly interesting: Can the computer create something? At first glance it seems obvious that it can. Animated computer graphics, with their fluid transitions and whiplash perspectives, look strikingly new. And if one watches the machine doing animation work, there seem to be lengthy periods when the computer is acting “on its own.” But if one observes these processes in more detail, it becomes clear that creation is not occurring within the machine. First of all, computer graphics are not unique. Computers have yet to generate anything that cannot be done by hand—and usually already has been done. Second, the apparent ability of the computer to “act on its own” is the outcome of thousands of hours of patient human effort to refine its instructions. The computer can manipulate a shape for us if we have already informed it what a shape is, what the rules for shape manipulation are, what this specific shape is, and so forth. You can start an automobile engine and it will run by itself, too, but that doesn’t mean it’s being creative. It’s just running.

Somebody in 1975 had a better understanding of why artificial intelligence is not in any way “intelligence” than the majority of today’s intellectual minds.

This is from a Tumblr post which I’ll link at the end and it’s food for thought. Because the way the general public think about Generative AI LLM’S is absolutely being spun by the people trying to sell it to the masses. That’s why arguments like this don’t get any traction.

https://www.tumblr.com/smol-blue-bird/718439525090803712?source=share

So you noticed, right? The point isn’t the predictive nature of intelligence, but keeping options open. So when a system is open in this way, it keeps its own options open and optimizes to create more options over time. So if we assume that intelligence isn’t about being right, but about keeping options open to escape being trapped or closed in.

This is a very good observation.

I think this is a characteristic of real intelligence because it is a poperty that is helpful to living beings (animals and anything that has some agency): Having options - for example, ways to escape from a potential predator, or ways to find different food sources - helps them to stay alive. And agency (not the LLM type) helps them to stay alive.

I think absolutely that real intelligence is very much a property of living beings. Because it has evolved to protect them being alive.

And perhaps because having agency also implies some amount of freedom, this might be why tech fascists hate living beings so much: Ultimately, they can’t control them. Tech fascism is an ideology of control, and the whole AI madness is an aspect of this. An attempt to make life and living minds non-special and replace them with something that is easier to control. And telling people that there is no alternative, no escape. It is a mental jail.

Sorry this probably isn’t the kind of engagement you’re looking for, but I absolutely HATE the way the article text keeps swapping between ‘AI’ and ‘LLM’ in real time.

I get what you’re trying to do, but every time the text swaps it forces the whole paragraph to realign and it makes me lose my place while reading. It also made me think I was going crazy the first time it happened, which is maybe not the vibe you want in an article about intelligence.