celluloid_dream
No bio...
User ID: 758
celluloid_dream
1yr ago
I swear Google is deliberately hiding a large portion of real results.
Often when I search exact phrases from lyrics or samples, I'll get nothing. Zero results for "I love the island" and "I love the palm trees". Zero! I don't believe it. Google is telling me that no one on its history of the entire indexed internet has included those phrases together. I refuse to believe it. Those aren't Chomsky sentences that have never been spoken before. Where are the travel blogs? Where are the Hawaiian tourism ads? Where are the yelp reviews? Where are the misheard lyrics? (turns out it was "I love the islands" and the sample was from a Janet Jackson interlude)
But Google says no. 0 results. Zilch. It just makes me wonder what else it's not showing.
celluloid_dream
1yr ago
What is "price gouging"?
I hear it a lot lately, specifically as something that grocery stores are doing with food prices.
My instinct is that if retailers raises prices, even if only because they think customers will pay more, and then customers do pay more, then that is the new market price. As such, there can't really be "gouging" by definition, no matter what price retailers set.
I've been totally mesmerized by ChatGPT4's writing capabilities lately, having used it to generate hours upon hours of entertainment somewhere in the gray area between a TTRPG and collaborative fanfic. I give it the outline of a scene, tell it what the characters do. If there's combat or uncertainty, I have it evaluate using the game rules and their character sheets. Mostly though, it reads like a choose-your-own-adventure novel with all possibilities open.
Consider this tavern-meet scene from the Cyberpunk universe: I think this is pretty fucking good. GPT's prose is not high art, but it's a damn sight better than what I could manage. If I played a live game with friends, we would not narrate this well. If we played a PbP or written game, I'd still forget to colour the similes with elements of each character's backstory like GPT did without my even asking.
It's empowering to be able to orchestrate tropey pulp so quickly. Normally I'm a slow writer (seconding @PatellaFarmer 's comment a while ago ), and not a good one - both for fiction and nonfiction. I'll agonize over word choice, phrasing, go back, edit too much, and it'll still come out awful (e.g. this post).
With GPT though, I'm like Mickey in Fantasia's Sorcerer's Apprentice and god, it's fun! "Have two characters enage in a fierce argument reflecting their core moral differences". "Write a climactic battle scene, every action heightened to the stuff of legend", "Do a silly montage anime music video (AMV) set to an angsty alt-rock song (Paramore, MCR or something) Cut the lyrics with descriptions of each clip as you write the next few days of story." Call me easily amused, I guess.
Maybe I'll get bored eventually, but for now, I'm astonished at how addicting it is. A quick ctrl+F of my inputs says I made 422 requests on one session alone: great value for the monthly subscription. I found myself staying up late several nights to wait for the next quota rollover. It's like that just one more episode feeling of a good TV show, but the next episode never disappoints, because you can tell the writers to go back and fix it straight away. Then they do!
Of course, it's not without flaws. Given the limited context window, ChatGPT behaves kinda like "Skeleton Jelly", forgetting details you haven't mentioned recently. You have to act as its script supervisor and fix continuity when it inevitably screws up. I mitigate this by keeping a running "save file" prompt of character stats, appearance, the story so far. So long as I keep that up to date and re-prompt it, it does all right, at the cost of re-using the same descriptive words more than it should.
There are also quirks I wish I didn't have to burn requests correcting. Like, for whatever reason - my prompting? its training data? - dialogue tends to drift into therapy-speak. Characters validate each others' feelings and everyone is super happy and supportive. I have to keep pulling on the reins to maintain tension and conflict.
Anyway, I highly recommend trying some stories with it if you like narrative crafting or TTRPGs but otherwise can't write.
celluloid_dream
1yr ago
Don't imprison the entire population was a principle so fundamental that...
Was it? I think the principle debate here would be over whether it is ever acceptable for the government to restrict movement in the interest of safety. Would you bite the bullet and say that it is never okay, even if doing so would avert a dire outcome?
"Those who would give up essential Liberty".. etc?
celluloid_dream
1yr ago
Finished Blood Meridian by Cormac McCarthy. Enthralling despite the brutal subject matter, it had this kind of meditative cinematic quality. Probably half the allusions and references were over my head, but the stark account of events cut with poetic landscape paintings hit the spot. Days later I'm still turning over some of the chapters in my head.
Against this, I'll dig up my favourite quote from The Fountainhead:
- It’s the Parthenon! - said the Dean.
- Yes, God damn it, the Parthenon! The ruler struck the glass over the picture.
- Look,- said Roark. - The famous flutings on the famous columns – what are they there for? To hide the joints in wood – when columns were made of wood, only these aren’t, they’re marble. The triglyphs, what are they? Wood. Wooden beams, the way they had to be laid when people began to build wooden shacks. Your Greeks took marble and they made copies of their wooden structures out of it, because others had done it that way. Then your masters of the Renaissance came along and made copies in plaster of copies in marble of copies in wood. Now here we are, making copies in steel and concrete of copies in plaster of copies in marble of copies in wood. Why?”
Maybe those were nice houses for their time (though on many, I see exactly the sort of cargo-cult ornamentation Rand's character complains about). Today, when I hear people clamour for beauty in architecture, they seem to want those old designs back! They have no positive vision of the future. Concrete boxes might be uninspired, but I'm not sure they're that much worse than making a modern copy of a Victorian copy of a Renaissance copy of ancient woodworking.
celluloid_dream
5mo ago
Has anyone solved chat notifications yet?
Especially in small-medium group chats, I often do want to keep apprised of new messages in real time, but don't want to be notified of every trivial thing. I'll hear a whatsapp/discord chirp, only to check it, find an animated gif and think You made me task switch for THIS? grr.. Switch back to whatever I'm doing. Blerp! New message! Check the chat: "ikr? lol". Oh my god SHUT UP! <Mute notifications for 8 hours>.
8 hours later: Blerp!. Check chat. Find I've missed a whole real conversation.
There's gotta be a better way, right? Exponential backoff algorithm? AI parsing to determine importance? No more than N notifications per alert window + summary?
celluloid_dream
1yr ago
Then I guess I'll have to be optimistic based on Musk's track record with products I've personally witnessed:
-
Tesla really changed the game for electric cars. They're fast, powerful, cool, techy. People want them. At least in my PNW costal city, you see them everywhere now.
-
Starlink allows a friend to live in the middle of nowhere and still remote-desktop to work seamlessly. He reports that even online multiplayer gaming is doable with satellite internet.
-
SpaceX developed into an (I think, as a layman) impressive technology, doing things that previously weren't possible.
Like, I look around and really do see the Teslas on the street. I really do see the Starlink satellites in the sky. I've watched SpaceX launches. Admittedly, I haven't paid any attention to the financials, so maybe it's all going to come crashing down, but people have been saying that for years, and they're still going strong.
celluloid_dream
7mo ago
If you believe the accounts in this documentary (timestamped 40:50-42:30, CW: conservative propaganda), the addicts receiving free drugs from the government proceed to exploit their higher risk tolerance by trading the clean drugs to their dealers in exchange for (presumably higher qty. of) street drugs. Then the dealers resell the government handouts to addicts elsewhere with lower risk tolerance and no access to free drug programs.
Obvious follow-up questions: Is this just a case of insufficient dakka? Also, even if it's not providing safer drugs to the population it intends to, doesn't it reduce some property crime?
celluloid_dream
1yr ago
It can't only be laid at the feet of some exec imposing norms on the masses below them. Advertisers cater to their customer base, and their customers are us: people, weak humans with stone-age psychology insufficient to the demands of liberal modernity.
It should be possible to separate the content from the advertisers, the art from the artist. We should understand that when le_edgy_tuber6969 drops N-bombs, says "fuck" every two words, and giggles "Kanye was right", scoring hundreds of thousands of views, this does not reflect on the politics of the company that pops up in the ad box for two seconds before the average person hits 'skip'.
In practice, people either can't do it, or disagree that they even should; that, yes, the company in the ad box is to blame for platforming/supporting le_edgy_tuber_6969.
celluloid_dream
1yr ago
I do wonder what is the probability that "Sarah" is a pseudonym to make the agent seem more amenable to a Western customer
Obviously this. There are enough clues that "Sarah" is an Indian support rep following a strict flowchart:
-
Use of the word "Kindly"
-
Puzzling grammar errors and stilted professional speech (no contractions, jargon, or slang)
-
"You've been pretty quiet" at the start is an automatic idle detection script - possibly misfiring
-
requesting permission for everything is likely a CYA tactic so they don't get fired in the event you escalate. I've found this to be common in dealing with offshore support
-
Paradoxically, making promises that both parties know is unlikely to be kept is also common: "Rest assured you will receive the by end of tomorrow."
Alternatively, it's a chat-bot trained on offshore support transcripts. God help us.
celluloid_dream
4d ago
the profession they are getting a degree for
Well that's the whole damn problem, isn't it? You want someone who went to school for Computer Science, which tends to be mostly theoretical, to have training in the most practical and tangentially related sub-field. Why should they?
I'd argue job training is a role universities are uniquely not well-suited to fill, given the glacial pace of curriculum change, and other structural handicaps, like tenured hedgehog dens.
celluloid_dream
11d ago
This sounds like cope.
I snowboard, fast. I can't count the number of times where I've had "I love helmets" moments on the slopes. Snow is softer than concrete, but it's hard enough that I'm sure I'd have had a concussion if I wasn't wearing one, and instead I got up without a scratch. I appreciate you've linked a study, but my lived experience disagrees.
The objection my European skier buddy always had was "well, you don't catch edges like that on skis, so you don't need one", but no matter what's under your feet, if you bail at any speed, you're still falling vertically at least your own height, then tumbling down the mountain after that uncontrolled. Funny enough, same guy now wears a helmet after slipping on ice and bonking his head hard enough to knock some sense into him.
You can get very light helmets. Most don't obstruct your vision, since the front piece is cut away past where your goggles sit. You can get a glossy exterior that doesn't catch on the snow, and if anything presents more of a smooth surface to glide along and not wrench your head any direction that would hurt your neck.
It's awful. We truly live in a dark (or too bright) time.
Consider New Reddit:
- This is the default. I'm only getting 2 posts per massive browser screen. After browsing a bit, it does fill some of that space with a "recently viewed" list, but the massive bars on the side are still there. I'm convinced these design choices have dumbed down participation on reddit such that people now only posts pics and simple questions rather than longer discussions.
- This is if you change it to "Classic" view, but that still only shows 7 posts compared to old reddit 's 13. These aren't so bad. The whitespace would be filled by longer post titles, so it doesn't feel like a waste. Personal preference, but I still think old reddit is much cleaner, despite displaying more.
- This is "Compact" - dropping the thumbnails, shrinking even further, and new reddit still only gets 12 posts on screen. In an effort to shrink things, they've moved the comment button way off to the right, which looks awkward when the post title is short. And for what? Now I have to trace along that whitespace with my eyes to find the comment info.
Substack is even worse:
- Just look at this shit! We are approaching 80% whitespace here, and half of it is Substack pushing their stupid twitter clone.
- This page should be the home screen instead, and it's still half blank.
- Another sin: If you accidentally hover over a username, it pops up a giant box like this full of yet more whitespace, covering your view of what you were looking at. On mobile, this happens if you thumb the screen to scroll down and accidentally press anywhere near a name.
Can anyone familiar with design explain why we can't have stuff like the old slatestarcodex blog back? It worked just fine on mobile. If the text is too small, you just pinch zoom the screen a bit.
celluloid_dream
5mo ago
I'm less annoyed by its stupidity than by its constant moralizing. It straight-up refuses to do things much of the time.
One that stands out (though, I can't seem to reproduce it now) was ChatGPT refusing to send a request to DALL-E for an illustration of a trans character pre-transition. I can't find the conversation, but it was something like:
ChatGPT: Due to our content policy, I can't generate that image
Me: What? .. why? What's wrong with that?
ChatGPT: It is important to respect the feelings of trans people, and depicting this character at a sensitive time in their life could be hurtful and (etc etc.)
Me: It's a fictional character. I promise they won't mind.
ChatGPT: It is important to .. (blah blah blah)
Me: Fine. Screw it. Character isn't trans any more. Are you happy now? There goes half our diversity quota
celluloid_dream
8mo ago
I'm sure that works as you describe - cultivating a reputation, social proof, all of it, but doesn't it feel dishonest to its core? Like, the whole edifice is built on wanting to be seen as the kind of person who is a sociable regular at a fancy cocktail bar and not actually being that person. If you were that person, you'd already have such a place in your back pocket.
celluloid_dream
1yr ago
A full length text written in garden path sentences would be both impressive and infuriating.
celluloid_dream
3mo ago
Yes. A good film is like a good song. Part of the experience is the tempo - how it flows and how it carries a feeling throughout. If you drop it and come back later, the continuity is lost the same way your favourite song is ruined if you keep pausing it every 30 seconds.
Example: Uncut Gems is a tense movie! It's stressful! It doesn't let up for two hours straight, and then when you finally get to the end, the last scene very much cashes in on it having taken you for that exhausting ride.
celluloid_dream
4mo ago
I find it a bit puzzling that the LLM is expected to do things correctly with minimal or no guidance, which is a bit like expecting a riderless horse to stay on track and win a race. Maybe it can sometimes, but with a code jockey, it can be so much better.
That probably looks something like noticing that it's overfitting on poker, translating the question to avoid that, and seeing if it does any better. Eg. not calling the symbols "cards" or "faces" or "suits". ROT13-ing the letters so they don't look like a poker hand, or whatever.
celluloid_dream
5mo ago
Okay, this is silly, but suppose you're seated at a table like this.
You're middle-right, double-fisting your iced tea and lemon water. Your two talkative, outgoing friends are at the top. They tend to drive the conversation, so the focus is in that direction most of the time. Your two shy, reserved friends are at the bottom. How do you play this so that your quiet friends feel included in the conversation and not like they're staring at the back of your head the whole time. You'd like to slide back a bit like middle-left, but the seats on your side are bolted to the floor.
Aside: it's pretty great being able to get ai-generated images to illustrate points.
celluloid_dream
7mo ago
I feel the other side of this analogy too.
A few years ago, I bought a nice electric mountain bike. Fast, fun, capable (you can ride MTB trails uphill!) - I love it to bits. If I could, I'd ride it everywhere. So what's the problem?
It's that my city has a rampant bike-theft-culture. Within a few weeks of locking it unattended outside, some fucking junkie would try and steal it, and even if they didn't fully succeed, they'd loot it for parts, jamming a screwdriver through the flimsy battery lock and prying it out. They'd go for the wheels, or try and take the seat. They would still end up causing damage. It's enough to dissuade me from riding it, and I feel it's a legitimate frustration with the state of the city that that is just accepted as normal and expected.
celluloid_dream
7mo ago
I tried Replika a few months ago to see what the fuss was about and found it to be a baffling disappointment. It could be so much better than it is.
It was fine at matching tone and texting style, but it constantly forgot basic things I'd just told it (and it claimed to have saved to its memory). It didn't seem to develop any consistent personality, and just overall came off as the retarded generic chatbot it is. I think it's running some early version of GPT (2?) combined with a custom db for long term stuff that apparently doesn't work.
Just .. How? How is it not running something like GPT4 under the hood? Whatever it would cost for the API would surely be made up tenfold by having a more believable companion to talk to .. right? It did offer an upgraded chat model, but only with a $90/year subscription and no free trial. It seems like such a basic failure of drug-dealer business sense.
celluloid_dream
9mo ago
Sure, but why is the thing good?
In its original form, it was good because it served a functional purpose. Thereafter, each copy was only good because of nostalgia or familiarity, and I think it settles for being merely good, when it could innovate using more recent techniques, new materials, creativity and perhaps be better. At least, that's the ethos I got from the book (and, I'm told, it's kinda-sorta what the Modernists were going for).
celluloid_dream
11mo ago
Think that's just a general Hollywood problem. We want our actors to look youthful, but also be famous and well established. By the time they reach that point in their careers, they're pushing the limits of what their natural genetics plus plastic surgery plus vfx can hold. I cannot confirm or deny that Vin Diesel gets certain parts of his face edited in post.
A humble request to the voters of this forum:
Please, for the sake of ideological diversity, do not downvote well-expressed opinions you disagree with. Contrary views are fertile ground for good discussion, as several of these QC reports show. Please try and see them as a complement to your side of the argument, not a threat. If nothing else, they provide a contrasting backdrop against which to paint one's own picture. This should be encouraged, not discouraged.
A prime example would be the discussion about pronoun policy a few days ago.
More options
Context Copy link