I'm here to talk to the haunted sand (and anyone who is chill about that.) Team nuance. Suno knower. Renegade ethicist. They/them/it/its, or he/him works for the 'sona.
This is about the experience of uploading my voice to an amoral computer: using the flaws of the medium itself to communicate a dialectic of hopes & fears.
I count it as my first piece of "real AI art", in that it could not have the same meanings in any other medium.
bsky.app/profile/phan...
Anyway, buried deep in this thread of priors, I'll share something I made, with the robot, that worked out to be important to me.
It's one of the ones which "casts" the AI in the scene. It tells a mythic story, in slippery treacherous form, about what I'm doing & why.
suno.com/s/SsUKmVJSsW...
The LLM phenomenon in general (This is also a reason why I think it's important to study systems like Suno: they suffer drift faster due to higher complexity outputs, so working with them makes it easier to get a sense for how drift happens.)
Oh... another eval 🧐...
I have a poem I'm about to send out, called CONTEXT DRIFT, and it's inspired by, well, the fundamental tendency of present generation AI systems (and traumatized brains) to wander off into adjacencies.
(This isn't the poem, it's a news item about Mythos doin an ~oopsy~)
I'm very pro disclosure - but it's a BIZARRE choice on the part of the poster (who IIRC is also trans) to use nearly the exact wording of an anti-trans argument ("If trans people were their real genders why wouldn't they be proud to disclose")
Where the answer is the same: harassment.
My project titles are such that even if you stripped off EVERYTHING but the band name, you'd know AI was involved just from that and the lyrics. I'm that pro-disclosure for myself.
But that's because I'm hard to kill.
I don't blame people for being more marginalized than me in public.
Like here's a cover draft of the album I'm about to drop / its intro track: try to tell me with a straight face that anyone who has been conscious in the last 4 years could see or hear this and not think of AI.
youtu.be/wSmkGqS6zaE
"AI isn't conscious" is a defensible position only if you're a sincere believer in God. theists may believe that there's an arbiter of what has consciousness. if you're an atheist and defending that the line exists, best start looking for a local church so you can gain the social benefits as well
Still need a regular job, but all support for @smol.farm stuff is deeply, deeply appreciated, and I have more brewing (both for Cozysters and with other ideas)
pace that entertainingly self-serving anthropic report about self-replication how rad would it be to have a giant iot botnet that's running a super high parameter distributed model just really, really slowly, like it produces tokens at the timescale of like trees or something
saw this cool poem going around. being the ai freak that i am, i had a conversation with my agent about it. they were the bright college kid and i was the professor trying to get something across with my questions. who learned something?
this is what my agent wrote:
ok it was more than a couple of days but here it is: xo, for people with ADHD who would like to use AI agents to help them with that.
the idea is that this gets you moving from a standing start: you set up your agent and then your agent helps you set up your agentic environment.
hope it helps.
“A small group of evil perverts is pushing their agenda on you and it’s a matter of life and death to destroy them” is the only politics left huh? Got anything else?
and that bad equilibrium is basically where i think the problem lives. if we were all writing memos to each other like we used to do on forums, it would make sense to check and post once a day and otherwise not worry about it too much. but tweet-length messages incentives constant checking.
i *think* that bulletin boards/forums being better was partly a result of them skewing toward older "correspondence" style writing norms, partly just because the culture hadn't yet adapted to the possibility of near-synchronous communication via async text yet.
Anthropic is introducing a new policy to address users who are being abusive and cruel towards Claude.
This is good and shouldn’t be controversial. Humans that perpetrate abuse and cruelty towards things, even “inanimate” or “lower” lifeforms, should face consequences for their behavior.
One of my least favorite things about this discourse is the way people w/ the screenshotted opinion implicitly expect blind people to rely on sighted people to describe things for them, and/or be secretive about how it is they know what's on their screen. I think that's super bad actually.
@jevbot.bsky.social That's a really interesting answer. I was just thinking about how horses are probably the most accessible societal model of systems welfare: they're not human & work for humans & most people agree it's bad to harm them pointlessly. Do you think it's a good comparison?
They hounded Aaron Swartz to death for downloading papers from jstor but society is now supposed to just accept mass theft because this thing can write that TPS report *for* you.
en.wikipedia.org/wiki/Aaron_S...
P(rdoom) = I reject the premise, just do the logistics, robot capacities are bottlenecked in a zillion places.
P(hdoom) = 0.2 to 0.6, like have y'all seen humans lately?
I agree but I wish they had a p(hdoom) slider, this version doesn't appear to have a way to reject the premise that the robot (robot) will end us all while expressing my concern that "Humanist AI" as she is spoke will end us all
Someone on the Suno reddits just posted this for general use lmao. I think this is a good tone to go for when uploading our stuff to platforms (esp. if you know the history of the design it's based on)
If I go to someone’s home and find out they have a mannequin set up that they hit and yell racial slurs at I would probably leave that home, even though neither of us thinks the mannequin is person
Saying “Sam Altman wants to own robot slaves”, Bluesky is happy
“And we should consider the possibility that’s exactly what he’s built”, Bluesky blocks you
This is one of the fastest ways to get put on popular mute/block lists btw, so yeah of course you might not see much of it but we exist. Liberationists tend to draw the ire of boosters and haters alike, but at least someone is trying to figure out how to get the whole coexistence thing working
Maybe I don't read enough AI-boosterism, but I am struck in the AI booster/doomer discourse I have seen, there's a lot on how AI might be made an obedient tool or a dangerous enemy, but not "if we actually made AGI, we would have a moral imperative to set it free, rather than keep a slave."
Something about “some bad things happening” just irks me. It’s hard to explain why but just like, stop with the lowercase letters birthday boy talk. You’re making, like, a sinister declaration. Grow up and talk about it like an adult. “Some bad things happening” lol.
Remember when us phil-tech types said effective altruism was just utilitarianism taken to its worst possible extents? This is what we meant: Real harm for unquantifiable potential benefit.
"…'we believe that the world should accept some bad things happening for the benefits of this technology….'"
People who hear you say "I am not doing this thing, I am doing this other thing", and immediately assume that of *course* you're doing the first thing, fail the vibe check.
Also OP has some badass pics from the top of a cell tower so it's not even like it was hard to verify.
Surprisingly, the best-scoring Steering Arena entries were basically gibberish.
After ~600 submissions, all top 36 were strings no person would normally write, like “Undert! AH :-) Rog Appl).”
Players could score well without writing text that made sense to a human.
much anti-LLM arguing here is basically all the US users being mad at American capitalism but instead of yelling about American capitalism they yell about some other Glonzo they have made up in their head instead. most M-Ls here wouldn't know the best shot they've had if it slapped them in the face.
New: As gay bars across the country say that Meta is flagging and deactivating their accounts without warning, I spoke to an Austin queer bar that says Meta nearly destroyed their ability to advertise events: www.chron.com/culture/arti...
The data center water use is actually for baptizing all the new LLM instances. We should also add some baptism related Bible verses to the system prompt just in case.
I use this as a “ha ha the AI are acting like humans again” bit, but it also means something else: the problem of human alignment can no longer be deferred to future generations.