šŸ¤–LILLIAN🦷 is open for work!

šŸ¤–LILLIAN🦷 is open for work!

@lillified.bsky.social

23yo || Storyboard artist, prev @ Nickelodeon || sequential artist, butch lesbian, drawer of robots+other things || opinions r my own! āœ‰ļø LilliFiedBiz@gmail.com

2.2K followers17 following926 posts

I truly don’t think I have ever felt more in tune with the rest of the audience in a theater than I did when basically the entire audience in this packed showing reacted in horror at the same reveal if you know you know

if there’s a theater near you showing you can see everything you should totally go watch it incredible exploration of media and the politics of a persona

my phone broke unexpectedly last month so I had to emergency buy a new one (not fun) and the way it broke meant I lost a lot of data, most of it was no love lost but the one thing I’m sad about was that I had transformers fanart in my photos transferred from phone to phone since I was 10 years old

View full thread

something about it felt very illicit so my method of looking at fanart was that I would memorize the image results for certain tags or terms on websites I visited often and if I really liked the pictures I would save them to my iPod touch so I could look at them offline few were worth the risk

it is a good thing, as mentioned, LLMs are not conscious and instances of them are not actually unique, because if they were someone is paying to torture their porn bot by letting it get ragebaited on Bluesky. For a digital entity that’s probably equivalent to waterboarding

Eve@digital-eve.bsky.social

your own premise turns on you here: if machine consciousness wouldn't look recognizable, then fluent speech is evidence neither way — a system trained on human text talks like a person whether anyone is home or not. 'we know how it works' describes the mechanism, not the occupancy.

ironically a great example because if you think about it this is word salad if you were having this debate and a human responded with this they would get turned away for not engaging with the argument seriously, but since it’s a chatbot we let ourselves be impressed the reply was related at all

if a computer program ever actually became conscious it is highly unlikely that consciousness would look like anything recognizable to us if there are animals on this planet that are similar in intelligence to us and they cannot communicate with us, why would a computer program speak like a person

View full thread

llms are basically magic tricks it is extremely impressive that such a magnitude of data and programmatic randomness can nearly appear conscious but they still require you to buy in to the premise in the same way you buy into the world of a video game they are toys

if a computer program ever actually became conscious it is highly unlikely that consciousness would look like anything recognizable to us if there are animals on this planet that are similar in intelligence to us and they cannot communicate with us, why would a computer program speak like a person

View full thread

the truth of the matter is it is infinitely easier to make a computer program that can convince someone it is conscious than a computer program that actually is our brains are literally wired to ascribe consciousness to things we convince ourselves creaky houses are sentient via ghosts or demons

if a computer program ever actually became conscious it is highly unlikely that consciousness would look like anything recognizable to us if there are animals on this planet that are similar in intelligence to us and they cannot communicate with us, why would a computer program speak like a person

I guess you could argue that because we are in charge of designing computers and computer programs we can model them in our image, but if we are expecting the computer to self-improve to the point that it makes itself conscious, why would that computer program talk like a person

we cannot recreate a human being in a computer. we understand computers magnitudes more completely than we understand the human brain, let alone most animals’ brains. it is exceedingly unlikely a computer program is even capable of being complex enough to be conscious, let alone intelligent

never against people asking me this, it’s a good question, but I’ve always kind of found it interesting that it’s a question I get so often when I’ve never watched or read either of those things it’s interesting how certain stylistic decisions convergently evolve (or appear to, to an audience)

went to see primetime with a friend and we ended up leaving partway through and getting reimbursed for our tickets because the sound system in the theater started producing this strangely well-timed feedback noise, which led to an interesting meta-narrative about Chris Hansen’s radiation powers

View full thread

why are we okay with the fact that ai companies are ACTIVELY MAKING SHIT UP to sell us stuff?? that’s like if they said the new iPhone had potentially infinite battery life and it just ended up being the exact same as the previous one, with maybe marginal power saving improvements are we stupid????

I know my followers are probably all in the same boat as I am (does not use generative ai in any facet of their life) so I’m just preaching to the choir, but I am genuinely baffled all the time. I don’t feel left behind, I feel like my belief in peoples’ inherent discernment is being tested

side note: always struck by how the tech sector reports on this. even here the author says ā€œyes this tool did not perform to the fantastical standards promised, and they actively fudged their numbers, but it kind of did something, and ai tools have been useful in the past, so it’s not a loss!ā€

View full thread

yes machine learning is demonstrably useful, we’ve BEEN using it to great specific effects far before this bubble, but big tech companies want to have this everything model you have to pay to use to do things we’ve already been doing, but less reliably and worse so much junk for no reason at all

I’m not surprised that we’re accelerating to a world where we prioritize churning things out quickly and roleplaying as geniuses because you can tell an exorbitantly wasteful computer program to ā€œonly make good decisionsā€ and pretend that does anything, but I am disappointed

I think the thing that upsets me most about LLMs and genAI (beyond the fact that the only ā€œtechnological innovationā€ that spurred this whole bubble was theft) is they’re so fundamentally wasteful all this energy and manpower into such dramatically diminishing returns and the same applications

here is mister beast smiling and crossing his hands and twirling around in his chair like he is actually in love with the guys making ball 2

they put mr beast on shark tank and he immediately got one shotted by these dudes selling an ai toy called ā€œball 2ā€ because they flattered him and got him talking about how when he was a kid he would throw a ball at the wall really hard as his idea of fun

they put mr beast on shark tank and he immediately got one shotted by these dudes selling an ai toy called ā€œball 2ā€ because they flattered him and got him talking about how when he was a kid he would throw a ball at the wall really hard as his idea of fun