Or: strenuously avoiding a national ID system, because of how obviously that could be abused. And probably still hacked anyway.

It is possible to do such an ID system sensibly, fairly, and for considerable net good. But we wouldn’t.

Rate proposed community notes

Presumably “a house costs more than a chicken.”

I keep trying to manually write code that I’m proud of, but I can’t.

That’s normal. That predates LLMs. Honestly, that predates compilers.

Basically every aspect of LLMs has been overblown - positive and negative. They obviously have some utility… which the self-proclaimed haters will never acknowledge, and want you to feel bad about using. But the robot will never match Sam Altman’s cocaine fantasies. It is buck-wild that ‘what’s the next word?’ works anywhere near as well as it does.

When Microsoft was still pussyfooting around, training on Github’s GPL projects, and their model occasionally spit out entire stolen files, it all seemed pointless. Now, years later - there’s a guy on Youtube who built a camera that visualizes laser pulses in-flight, and halfway through the video he laughs and hand-waves the code for it. Another guy builds video-game aim-bots that physically move the mouse, the table, or his musculature, and similarly brushes off the part where the computer does stuff.

We have programs that write programs. That’s just a thing, now. If the bubble pops tomorrow, it’s not going anywhere, because local models will run on a Raspberry Pi. (Admittedly that example was an art project where the AI waxes poetic about mortality until it runs out of memory and reboots.) We’re in a stupid manic phase, but a decade from now, spicy autocomplete will be just another tool. F7 for spellcheck, F8 for grammar check, F9 for the Dixie Flatline to do his best at whatever you ask while he’s kinda drunk.

The answer remains, this tool is not flawed, it’s just not the one you want.

Vim could be feature-complete and formally verified and I’m still using Xed.

Notepad in Windows 7 occasionally did some weird shit.

You cannot criticize a good knife by asking why it’s not a hammer.

Didn’t help Bleem.

which is how they killed Ryujinx.

Plus sending goons to his house.

Allegedly.

If you think Democrats should be the worst we can do, we have to stop doing so much worse.

Advertising shits in your brain.

Let’s get rid of it.

Your business model is not my problem.

Especially when plenty of profitable services add this shit anyway.

Read: Discord is de-anonymizing all users.

Leeeave.

Consider this image. It’s full of blatant tells, like the bouncer becoming a real boy from the knees down, or the improbable stacking inside that trenchcoat. Yet it obviously conveys meaning in a clever way. You wouldn’t commend whoever made it for their drawing skills, but the image transmits an idea from their brain to yours.

The model did not have to comprehend anything. That’s the user’s job. A person used the tool’s ability to depict these visual elements, in order to communicate their own message.

If some guy spends days tweaking out the exact right combination of fifteen unforgivable fetishes, that amalgamation is his fault. You would not blame the computer for your immediate revulsion. It tried its best to draw a generic pretty lady in center frame. But that guy kept doodling a ball-gag onto Shrek until image-to-image got the leather strap right, and once he copy-pasted Frieren behind him, it just made her lighting match.

Neural networks are universal approximators, so you’re always going to need human art to approximate human art. However, there are efforts to produce models using only public-domain, explicitly licensed, and/or bespoke examples. (Part of the ‘do words matter’ attitude is that several outspoken critics scoff at that anyway. ‘Like that changes anything!’ They’ll complain about the source of the data, but when that’s addressed, they don’t actually care about the source of the data.)

Personally, though… I don’t have a problem with using whatever’s public. For properly published works, especially: so what if the chatbot read every book in the library? That’s what libraries are for. And for images, the more they use, the less each one matters. If you show a billion drawings to an eight-gig model then every image contributes eight bytes. The word “contributes” is eleven.

Recognizing that no tool is immune to human expression. So even if a stick-figure single prompt isn’t art, some weirdo pouring their time and energy into an iterative process should be.

Distinguishing capitalist implications of a technology vis-a-vis material impact on existing professions, versus people running some jumped-up chatbot and renderer on their own desktop for their own purposes.

Frankly it should be illegal.

Not just “never required.” Explicitly disallowed.

I feel I’ve been pretty respectful throughout this conversation.

I wish to disabuse you of that notion. Every response has been condescending repetition of your apparently dogmatic opinions - and insistence upon your right to hold them, as if anyone challenged that. Conversations are supposed to pursue a mutual understanding of reality, instead of spitting conclusions at one another. You can’t make declarations about all art with a sweep of your hand, and then flinch in confusion when the person you’re talking at has a follow-up question.

You “don’t feel the need to convince you or anyone of anything,” but boy howdy you sure keep yapping. And then cannot imagine why that’s not the end of it.

Meanwhile, I’ve held the vain hope this interaction might be productive in some way. From the very start, I asked you: do words matter? And you clutched pearls as if the answer was obviously yes. But then ev-er-y sin-gle response, that one included, ends with useless ‘agree to disagree’ fluff, and some sign-off like you’re just going to nope out, and then you keep coming back to do it again. I’ve made it crystal clear why I’m still trying. I for one give a shit about this topic, enough to constructively discuss it. Why the fuck are you still here if you’re not even listening?

You can’t go from ‘it does harm! fact!’ to an appeal from the nature of beauty.

Nor can you honestly claim ‘this is an open forum for debate’ and then make crystal fucking clear you’re not actually arguing, you’re just spouting words at someone. You came to me. I did not seek your opinion, I did not invite brainstorming, and I’m not keeping you here. But for some fucking reason every single response includes a pearl-clutching rejoinder like you can’t figure out why someone is talking back to you. All you did was offer concrete rationale and claims to fact which are supposedly relevant to the broad and important topic at hand! Why would anyone have something to say about that?

If addressing your alleged reasons cannot possibly change your mind, then they’re just lies. That’s pretense. That’s pulling out excuses that sound like arguments, but cannot be foundational to your conclusions, or else removing them would make that belief structure collapse.

If you don’t actually want to discuss this, consider shutting up.

At what point were you engaged in debate? You act bewildered that someone tried to critically examine your assertions. I’ve been asking questions from your opinion. Do you not recognize your own stated beliefs?

That is a fact, not opinion.

Incorrect. Locally running ComfyUI or Ollama neither picks their pockets nor breaks their bones. When these cloud companies crash, local models aren’t going anywhere, and they won’t do psychic damage to someone whose DeviantArt posts were in the training data.

Would any difference there be specifically attributed to any individual chatbot prompter?

In one shot, probably not. After a week of fucking around to pursue a specific idea in each person’s head, almost certainly. Then again, you could have one group of people prompt a hundred images total, and a completely different group each pick their favorite ten, and that non-interactive selection would reveal individual aesthetic internality. Curation is not creation, but it requires identifying a work’s shortcomings, even if you won’t correct them yourself.

Would a demonstrable difference change your conclusion? Like, is this line of questioning relevant, or are we just saying words recreationally?

Then why are we talking?

You seem to think opinions are decorative things, having no bearing on reality. Like the topic at hand doesn’t practically impact entire career paths for millions of people. As if the only reason to try to dissuade someone from dismissive absolute rhetoric is if I am somehow swayed by it, and secretly believe the opposite of all the words I’m saying.

Solipsistic time vampire. Why does anyone give a shit where you stand, if you don’t know what it means to stand by it?

If that journey is multiple days of fucking with these tools, why does that not count? Why is this the only technology immune to human expression?

I don’t need the constant reminders of what an opinion is - but you might need a refresher on what arguments are. If you give a reason for an opinion, people will often assume that’s why you believe something, and address it in a way that may alter your conclusion.