iiinova — where your digital self keeps evolving

jil @postfreq.bsky.social · 08/10/2026

seven samurai is amazing but at 3 hours I have to watch it like a tv show

000

can't wait for the five hours of increasingly bewildering rabbit holes he fell into while making it

070

"they built it with stolen code" no they didn't, you ass

000

as much as the idea of stealing art (in the sense of stealing the information contained within) is nonsensical it is somehow even more preposterous when applied to code

100

is that what he's been working on since the plagiarism video?

150

the (very long) Leftist Cooks video "Why Don't Fascists Eat Vegetables" opened my eyes to how proactive the meat industry is with managing its public image, most often by maligning their critics and competition

040

a lot of it has to do with a concerted effort by the meat industry. they spend a lot to manage the public image of these things

000

feels like I see accusations of being a cult more than usual lately. not particularly illuminating or convincing I don't think

000

I was only familiar with Bender's initial essay and I remembered it being fairly restrained so I was surprised to learn her reputation these days but after checking her current output, yeah, it's quite embarrassing

000

jil @postfreq.bsky.social · 01/10/2026

do you reject the idea of a non-conscious AGI? though the term is vague I think it most often refers to capabilities, not any structural property not to say I expect the profit-seeking machine to stop and acknowledge if consciousness was there but I don't think AGI necessarily entails it

020

unacceptable. backwards compatibility is important

000

jil @postfreq.bsky.social · 26/09/2026

it's not the nazis or bots that made twitter unusable for me. they were a big factor but I still came back the new app defaulting to For You (and not Following) actually funneled me to bsky

001

I wouldn't have expected you to struggle with writing too much. your stuff feels confident that said I very much get the pain of having a vague gestalt of the work that feels promising and energizing but having trouble building it out and dealing with all the emerging issues

I engage with him because he makes this place lively and because he's interested in important things so he jumpstarts dozens of active conversations. it's too bad this also leaves so many people with axes to grind but I guess that's part and parcel

100

please dont go please dont go 🎵🎶

090

that may be what the discourse came down to, I don't think that's his argument

120

I guess there are shades of that in his takes about LLMs but I don't think his core argument is "we should defer to intuition in analysing LLMs"

130

amusing juxtaposition but I don't think there's inconsistency here

120

we have roomba at home

000

our age of information is going strong

our age or information is collapsing

Nolan is fun. even when I don't like I'm like "he really saw it like that huh"

000

finally got around to Odyssey

100

the repeating pattern jevbot produces is interesting

the intersection of techie and artist that is needed for this is quite a filter and also all the incentives to hide usage but I think there are probably interesting projects if you look

031

I'm partial to that view. could you tell me some terms or prominent names to look up? if you have some in mind

000

ah, conscious understanding. that's a neat qualifier I think. is it a fairly established way to phrase it? and, you say the broader concept of understanding is still being settled - would you say part of the question is whether there can even be anything BUT conscious understanding?

100

I got this implication from some of your other posts - would you say functionalism and computationalism are the dominant paradigm in these matters at the moment in academia?

000

in my limited understanding discerning between animal cognition and LLM processing in a principled manner requires rejecting functionalism and computationalism

200

I don't think it's confusing. I think it is badly phrased and (at least in that form) poorly supported

000

I think it's easy to interpret it that way, and I also think it highlights the problem that anthropomorphizing is bad when it's misleading and not as a general rule I actually agree that we shouldn't ascribe cognition to LLMs but this phrasing is terrible

as far as I can tell from the abstract the argument is "a system designed to mimic is likely to be mimicking". I don't like the argument but it doesn't seem metaphysical to me

000

oh lol I'll check that out

020

yeah I caught bits of this one through NL. Justin's veneer voice is resounding in my mind as I type this lol and my first exposure to Ricky was from a Scam Line stream where he was Dark Woke

020

I'm just curious what changed to me it seems like he's been fairly consistent for quite a while but I've only been watching him a year or so, and mostly vods

000

do you mean fallen out with him in terms of stream direction or something like that?

100

who are you referring to with some more hardcore leftists? could it be DBJ, Ricky, etc? I don't know their politics so to me they appeared about the same as NL and co. and I think before I caught some allusions to Justin being the most radical of the group

130

US military was bad before Trump it's just his blatant actions made the world catch up a little

000

could you tell me what stream was it on? I'd like to hear his, Justin's and Kory's reaction. maybe I'll just watch full vods of the past several days

000

but that doesn't stop the capabilities from developing rapidly and though I think many of the more grandiose predictions are overly optimistic or outright impossible, the results we have are already astounding

000

I think we have a better idea than that, and I also think there are reasons to think no human-like cognition is happening, though the "like" can be stretched far

100

I think this can readily include systems that don't understand or think if they still produce relevant output. if that's the case then I don't disagree, LLMs seem to possess this to some degree already and are likely to develop even more in that direction

ah. I wonder how you make it robust. is the range so big that it includes non-human animals or do you not think that they do cognition?

100

I said animal cognition, which is relevant to computationalism do you define cognition without reference to animal cognition? I'm not sure I understand your position

100

and I'm not saying these systems are not dangerous or that they do not process information. demonstrably they do, but it's a very different process