iiinova — where your digital self keeps evolving

Ashley @polyaletheia.bsky.social · 3h

Why does "Centers & Corridors" sound like a Minecraft update?

010

Ashley @polyaletheia.bsky.social · 3h

OK but compact disc data goes from the inside out.

000

Ashley @polyaletheia.bsky.social · 08/10/2026

"Mathematicians did not ask for this work to be done" lol. gatekeepers get what they deserve

060

Ashley @polyaletheia.bsky.social · 02/10/2026

In practice this is going to make TSA a less desirable employer, which will lead to understaffing (annoying travelers) and/or wage hikes (paid by the taxpayer).

000

Ashley @polyaletheia.bsky.social · 02/10/2026

But AI is just a techbro fad like crypto [checks Bitcoin price] er I mean NFTs!

040

Ashley @polyaletheia.bsky.social · 01/10/2026

The only time I've interacted with him he insisted on his own invented definitions for "playing chess", "solving chess", "cognition", "capability", "fascist", and "functionalism".

091

Ashley @polyaletheia.bsky.social · 25/09/2026

Jack Lynch makes a good point about responsibility. While it's absurd to complain about anthropomorphic language around something already somewhat human-shaped, I do think we should watch out for implicit assignment of responsibility to machines.

181

Ashley @polyaletheia.bsky.social · 24/09/2026

Your description doesn't include anything newsworthy. Probably lots of people "create autonomous hacking software and leave it running for hours and days with no one watching", so what? You didn't mention any actual harm.

250

Ashley @polyaletheia.bsky.social · 24/09/2026

That's 1.8 MJ/kg in SI units or about 20 pico square lights

010

Ashley @polyaletheia.bsky.social · 23/09/2026

No particular regulation; existing laws apply.

000

Ashley @polyaletheia.bsky.social · 23/09/2026

Capability is function. If you are talking about capability you are already within a functionalist frame. Therefore, it makes no sense to appeal to "non-functionalism" nor to claim it is an illusion. If it fulfills the function, then it genuinely has the capability.

100

Ashley @polyaletheia.bsky.social · 23/09/2026

Will: can we just talk about the capability for a moment? "Boxo McFoxo": that's fascism

210

Ashley @polyaletheia.bsky.social · 23/09/2026

Right. And Will has made it fairly clear that he's only making claims within a functionalist perspective, and setting aside metaphysics, consciousness, and so forth.

110

Ashley @polyaletheia.bsky.social · 23/09/2026

No, you've just invented a definition of "cognition" that requires consciousness. You'd make more sense if you stopped inventing your own definitions for words.

000

Ashley @polyaletheia.bsky.social · 23/09/2026

It's not specifically "my" definition of play. It's everyone's definition of play. And I believe you already know this.

000

Ashley @polyaletheia.bsky.social · 23/09/2026

Empirical benchmarks are evidence for this.

100

Ashley @polyaletheia.bsky.social · 23/09/2026

Will has made it very clear that he is not talking about consciousness. He is only talking about presented capabilities. But people are SO desperate to deny AI any serious capability that they try and make all sorts of irrelevant metaphysical arguments.

110

Ashley @polyaletheia.bsky.social · 23/09/2026

The only way to prevent the possibility of non-conscious cognisers is to invent (again) a definition of cognition that requires consciousness from that start. This is exactly the kind of semantic bullshit that Will is complaining about.

110

Ashley @polyaletheia.bsky.social · 23/09/2026

"the first ever non-conscious cogniser" is no more unlikely than "the first ever non-conscious chess player". Or even "the first ever non-conscious chess solver" if we have to use your bullshit invented terminology.

110

Ashley @polyaletheia.bsky.social · 23/09/2026

These are entirely invented definitions of "solve" and "play". A chess-playing computer plays chess. It's very straightforward. All it needs to do is play (not solve) a good game. It doesn't matter how it does it. Cognition is the same thing: all that matters is its presented capability.

110

Ashley @polyaletheia.bsky.social · 23/09/2026

A chess-playing human solves chess. They solve it well enough to beat another human, which is how their games are won. They don't need to solve it fully to beat another human. They just need to solve it: to some degree.

100

Ashley @polyaletheia.bsky.social · 23/09/2026

Yes, we have a widely-used term for "coming close enough to solving chess to beat a human" and "getting the job done", it's called "playing chess", and it's also what a human does if they're good enough.

100

Ashley @polyaletheia.bsky.social · 23/09/2026

You made up your own bullshit definition of "playing chess", and apparently you've also made up your own bullshit definition of "solving chess". Why would anyone take you seriously?

000

Ashley @polyaletheia.bsky.social · 23/09/2026

Chess has not even been solved! Go read this: en.wikipedia.org/wiki/Solving...

en.wikipedia.org

Solving chess - Wikipedia

210

Ashley @polyaletheia.bsky.social · 23/09/2026

You invented your own complete bullshit definition of "play chess" that no-one else uses, just to win an argument.

020

Ashley @polyaletheia.bsky.social · 23/09/2026

Playing chess does not require intent; you just pulled that notion out of your ass. Anyone will tell you that machines can play chess, using those words. And this kind of made-up semantic bullshit is exactly what Will is complaining about.

140

Ashley @polyaletheia.bsky.social · 23/09/2026

You can literally go play games of chess with machines. And they will beat you every single time.

120

Ashley @polyaletheia.bsky.social · 23/09/2026

The only chess players on the planet require consciousness for their chess playing to function. By claiming that the machine is a chess player, but not a conscious one, you are claiming it to be the first ever non-conscious chess player.

150

Ashley @polyaletheia.bsky.social · 23/09/2026

Yes, so what? But are they capable of credibly presenting emotion? That's the relevant "human-like capability": after all, everyone agrees that they are not actually human. Can they present themselves as human? With regards to intelligence, certainly.

040

Ashley @polyaletheia.bsky.social · 23/09/2026

Is this the capability to credibly present emotion, or the interior conscious experience of actually feeling an emotion? I would say AI probably already has the capability, just not the consciousness.

140

Ashley @polyaletheia.bsky.social · 23/09/2026

Maybe try actually specifying the dumb thing that Will said, so he can address it?

020

Ashley @polyaletheia.bsky.social · 23/09/2026

Did you have a human capability in mind, where AI falls short?

140

Ashley @polyaletheia.bsky.social · 22/09/2026

Furthermore this is the more ordinary sense of "intelligent", the one that everyone uses, or at least did until the AI denialists ran out of pitch to move their goalposts to.

010

Ashley @polyaletheia.bsky.social · 22/09/2026

The key difference is between phenomenology and metaphysics. From a phenomenological point of view, AI is intelligent because it presents to us as intelligent, regardless of the "underlying" metaphysics.

110

Ashley @polyaletheia.bsky.social · 22/09/2026

very positive very positive very positive EXTREMELY FUCKING NEGATIVE jfc somewhat negative

020

Ashley @polyaletheia.bsky.social · 22/09/2026

Yes, actually, and pretty much any other kind of expert advice too.

020

Ashley @polyaletheia.bsky.social · 22/09/2026

And medical advice, actually. It's not perfect, but then nor is a doctor of course. ChatGPT (for example) is good for "this thing happened, should I see a doctor" and "this is what my doctor told me, what do you think".

140

Ashley @polyaletheia.bsky.social · 21/09/2026

I'm squarely in the first position, however, I think it's important to note that one cannot get from there to any kind of "subjectivity" or "consciousness" or "qualia" or "interiority" or AI, or in short, any reason we should ethically care about their "wellbeing".

150

Ashley @polyaletheia.bsky.social · 21/09/2026

Illustration is anything AI can do competently (i.e., no-one can tell any difference in quality in a blind test); art is everything else.

000

Ashley @polyaletheia.bsky.social · 20/09/2026

Better principles (just loosely thought on a Sunday afternoon): 1. Approach phenomenologically, not metaphysically 2. Require experience 3. Avoid political fantasies 4. Accept demonstrated achievement 5. Accommodate rapid development

001

Ashley @polyaletheia.bsky.social · 20/09/2026

This is actually part of the problem. Expertise needs to be demonstrated to be relevant, not asserted.

210

Ashley @polyaletheia.bsky.social · 20/09/2026

Honestly don't think credentialism is helpful at this point.

140

Ashley @polyaletheia.bsky.social · 20/09/2026

Wait, is this the "Cosmism" in the supposed "TESCREAL"?

010

Ashley @polyaletheia.bsky.social · 20/09/2026

The only part of the love tech tree that ultimately matters (as in, even beyond death) is the breeding achievement and cishet folks, on balance, have some advantage in that.

101

Ashley @polyaletheia.bsky.social · 18/09/2026

Got rejected at the first question for being post-rationalist??

120

Ashley @polyaletheia.bsky.social · 17/09/2026

I regret to inform you that none of the curiosity on BlueSky is ever in good faith.

390