CapsAdmin

I'm of the opinion that AI represents the average creativity, which can be interesting (especially base models) but on the other hand, too many cooks can ruin a dish.

It's also somewhat considered lame and cliche to ask an artist where their idea came from, and what their intention was.

However people will claim that the artist had an intention, it was just not obvious to the artist at the time. The art was also shaped by their experiences, etc. So in contrast, AI does not have this intention.

Maybe we care about where something comes from, but we don't have the will to care about everything? (also, never meet your heroes comes to mind regarding the endless debate)

For example, I had never thought or cared where the trash can in my office came from. I forgot where I bought it from. Now that I'm thinking about it, I'm a little curious, but is it worth knowing? If I knew the designer of the trash can was a bad person, should I get get a new one and make sure the designer of that trash can is a good person?

I think logically and ethically yes, but I'm not willing to spend my time on this.

It's an extreme example, but I think the same applies to music, art, movies, etc, just to a lesser degree for people who claim they care more about the outcome than the human author.

I think most people care about the final output, especially the market. This is why many artists complain about mainstream art, because it genuinely isn't interesting from an artist's perspective.

I'm not too familiar with visual arts, but I know most jazz musicians make a living off of teaching music and doing concerts where the attendees are just other musicians. It's very much centered around technique and human performance. (the term musician's musician is also used here)

Meanwhile, the average person think jazz is noise.

So my point is, saying "What makes music meaningful..." sounds more like an elitist jazz musician take. (I'd say jazz musicians tend to be more self deprecating than elitist, and often make light of the genre and how people perceive it!)

In these discussions, the concept of perceiving content gets intertwined with how content was made, because artists and creatives are usually more interested in how something was made.

So unless these two concepts are separated, people will endlessly talk past each other assuming both sides hold the same fundamental beliefs.

Claiming you made something by yourself, when in reality someone else made it (AI) is easy to frown upon.

I think if you can have a positive experience from looking at a sunset, hearing birds sing, etc, it should technically be possible to have a positive experience from the outputs of an AI without human input. This assumes art is clearly defined as needing to be made by a human with some effort, which neither a sunset or AI outputs are.

In practice, people who use AI to make content give it some direction. The amount of guidance varies wildly, and in practice the majority of what we see is minimal involvement.

I think then it's pedantry. I can kind of see how not having an opinion about what to wear tommorow is by itself having an opinion on what to wear tommorow, but that's not what people usually mean when they say they don't have an opinion about something.

For certain viewpoints that are fundamental ly hard to challenge, a reasonable middle ground to me is to just talk about something else and respect other peoples viewpoints. This is usually how religion is (or should be) treated in open public spaces. But it's by no means easy and I can't really blame vegetarians for being passionate about it either. But I see this more from the perspective of how to behave around other people and not so much what I personally believe, but maybe that's largely cultural. (Norwegian)

For example, I have strong opinions about people believing in things without sufficient evidence, but unless I'm in the correct space or is invited to, I'd rather keep it to myself.

is happy with whoever rules whom and how justly

This is an assumption. How about neutral, or even hopeless? There's also the position of just not caring because you think it doesn't affect you.

It feels a bit like saying anyone who does not want to fight or talk about a war are on the enemies side.

You can turn off the non-technical comments by ignoring them.

This also feels like an odd thing to say. If you're going out with a bunch of friends to eat out, and half of them always talk about how immoral it is to eat meat amongst themselves while you're eating meat, I think that would affect me negatively at some point.

One interesting workflow I've seen is that the project maintainer simply rewrites and implements the pull request themselves and closes the PR.

LuaJIT has operated this way since 2012, though with a thanks and mention in the commit message. It seems like a good way to filter out people who prioritizes leveling up their github profiles.

Something a little bit similar, when I was hosting a social game server we had mods. And players always beg for mod status. At first I tried naming the admin group something weird like sandals, but eventually people would ask if they could be sandals too.

What worked best in the end was just hiding it completely making regular players see mods as other regular players. (mods would see who is a mod though)

I would also personally never make someone who asks a mod as it's almost always a sign of wanting power for the sake if it. I would instead just passively observe behavior until I trusted the player and make them a mod. I would then tell them that I don't expect them to exercise their power, but would demote if I see abuse of power.

I'd say it's objectively true to say that open source software is easier to audit compared to closed source software, which you can extrapolate to mean that it's less prone to malicious code injection.

It's not perfect, but surely it's easier to audit for malicious code than closed source.

Also, there is no shortage of volunteers looking out for code changes in established open source software. I think it's fair to exclude software that is very new and/or that has no users, which may be closer to equal footing with proprietary software.

Even for established proprietary software, you get volunteers watching out for changes in releases. Though, far less than open source, and more reserved for people who know reverse engineering.

This is my takeaway as well. Having the source code open makes it auditable, if not by you, maybe the community.

The free software license specifically gives the software an extra advantage in that changes to the software must be shared openly, if distributed as as binaries.

unless it has video input, i wonder if something based on animation and timing would work, as screenshots wouldn't clearly capture motion and response time would be too slow as well

I'm not anti-AI but something I've been thinking about is the discipline it requires. As you said, it's a tool that allows you to rename a variable name on one end and do complete vibe coding on the other end. Developers may say that we should stay somewhere left on that spectrum, because that's where human's are more involved.

But developers also say good practices should be followed when talking to each other, and while some may do, reality is often very different.

It requires discipline, which varies a lot between developers, between projects, current mood, and so on.

In the beginning you might be careful doing small changes, but after a while you might get more tempted to accept the output for what it is, because ultimately that's much easier.

So the way I see it; the left side is harder work and potentially bigger but delayed dopamine hits, the right side is quick dopamine hits. How do we (at least those who struggle with discipline) resist just slipping to the right?

I started out carefully myself and slipped more into vibe coding, but I don't feel particularly proud of it for some reason.

Just to steelman the GP; some people in the company made a choice while the rest had no say.

I personally think the owners should get to decide, but it's an interesting duality.

(assuming it's not like everyone has a share or something, in which case they would've all had to agree I guess)

It seems to me that since the advent of image generators, art has been firmly defined by artists to mean that it was made by a human. But there might be a spectrum of human involvement where the less a human is involved the less it's art.

What happens too often during these discussions is that someone who writes "make me a cool image" gets conflated with someone used ai to fixup a small rock in their natural landscape drawing. (two extreme ends)

One problem though, is that we don't really know how much the supposed human author was involved in the piece. Now that it's becoming hard to judge, people against ai art can proudly change their opinion on on a piece once they learn that it was made by ai. I've come to think this is somewhat respectable, like you see a video of some extraordinary event (before ai) and then you learn that it was fake, just for views or something.

But on top of all this, there are different ways to "consume" art. Artists may think more about who the artist is as a person and what they felt when they made the piece, while non-artists may just enjoy the piece for what it is, detached from the artist. These two perspectives clash a lot.

While this is all good practice in theory, I wonder how much discipline plays a role here?

I am not very disciplined, and find it too convenient to reach for an agent these days.

This may sound ridiculous, but I am addicted to nicotine. I used to have some sort of rule around how I am allowed to use nicotine pouches to manage my addiction. For example after I finish writing a feature, I could have one pouch. It was obviously a dumb idea that didn't last very long.. But in that specific aspect, coding agents feel similar. I tried setting up rules on how I should use them, but it's not easy to follow them.

Maybe the biggest problem is just guilt?

I think it's not too different in that specific sense, but it's more than that. To bring libraries on equal footing, imagine they were cloud only, had usage limits.

I'm also somewhat addicted to this stuff, and so for me it's high priority to evaluate open models I can run on my own hardware.

I'm starting to think I've been A/B tested, because this was my experience for almost a year with Claude ever since I tried it for coding. Meanwhile, my coworkers seemed to be able to use it for long periods of time without getting rate limited.

One interesting variable is that I'm located in Vietnam while my coworkers are located in Norway and Europe.

To work around this issue I used Claude for coding with a Copilot subscription which was much cheaper and had virtually no rate limiting.

Copilot gives you some set amount of credits each month, but you can also pay as you go if you run out of credit which is much better than the 5 hour window crap claude code would give me.

The only opus model available now on copilot for some reason is 4.7 and it costs 7.5x tokens, while everything else is 1x, 0.33x or free.

But I switched to using GPT 5.4 medium for a month or so which I find very reasonable.

I haven't tried 4.5 haiku much, but i was not impressed with previous haiku versions.

My goto proprietary model in copilot for general tasks is gemini 3 flash which is priced the same as haiku.

The qwen model is in my experience close to gemini 3 flash, but gemini flash is still better.

Maybe it's somewhat related to what we're using them for. In my case I'm mostly using llms to code Lua. One case is a typed luajit language and the other is a 3d luajit framework written entirely in luajit.

I forgot exactly how many tps i get with qwen, but with glm 4.7 flash which is really good (to be local) gets me 120tps and a 120k context.

Don't get me wrong, proprietary models are superior, but local models are getting really good AND useful for a lot of real work.

I pay for copilot to access anthropic, google and openai models.

Claude code always give me rate limits. Claude through copilot is a bit slow, but copilot has constant network request issues or something, but at least I don't get rate limited as often.

At least local models always work, is faster (50+ tps with qwen3.5 35b a4b on a 4090) and most importantly never hit a rate limit.

It kinda looks like employees need to make a blog post about something twice a month.

I moved from a very quiet culture to a very noisy one.

Here people watch tik tok on full blast, people let their kids run amok in concrete cafes, constantly honk at each other, blast karaoke for all neighbors to hear, etc.

These people have some ability to sift through noise. For example being able to talk to someone on the phone with a loudspeaker in a loud environment while both seem to understand each other well.

But for some reason, the majority of people don't care, and so in some weird way, the concept of sound pollution don't exist.

When sound pollution don't exist as a concept, there is nothing to get annoyed about.

If the definition of art is that a human must be involved, then fine. AI generated music is not art. But it is everything art is minus the human component? ie, it can be beautiful, ugly, etc, just like how a sunset can be beautiful and a rotting corpse can be ugly.

I appreciate the honesty. I'm not saying people don't have this relationship with art, I think everyone can have some degrees of it, including me.

But my experience as an artist talking to non-artists about art, I don't think the sentiment that art without a struggling artist, purpose, story to tell, human arc, etc, is not real art is a true sentiment. First of all, because it's not true, because people apply their own meaning and form their own unique relationship with an artist. (The saying don't meet your heroes come to mind.)

Note that I'm not talking about AI at all here. I'm 100% for banning purely generated AI on soundcloud, bandcamp, spotify, etc. What I really want is to filter out art created by people who has put profit as first priority and thrown away any shred of artistic integrity.

But this is an impossible feat, because who am I to judge that someone else's favorite artist is devoid of artistic integrity?

I think maybe we're talking past each other then. I'm saying I don't agree with the argument that music necessarily needs to have a story to be widely consumed in a positive way.

While I personally like it when people put their heart and soul into something, even if the result is technically not very great, it's society who is the ultimate judge of whether that creation benefits them or not.

I know that the track I'm currently listening to is superior in every way to some modern pop song. The artists have practiced for decades, they have their own unique style I can recognize in other tracks. But I also know that 99.999% of people don't give a shit and think it's noisy music, and depending on your perspective, they're correct.

This has been my experience as well. But, these are things we developers care about.

Coding aside, LLM's aren't very good at following nice practices in general unless explicitly prompted to. For example if you ask an LLM to create an error modal box from scratch, will it also implement the ability to select the text, or being able to ctrl c to copy the text, or perhaps a copy message button? Maybe this is a bad example, but they usually don't do things like this unless you explicitly ask them to. I don't personally care too much about this, but I think it's noteworthy in the context of lay people using LLM's to vibe code.

I've been playing with vibe coding a lot lately and I think in most cases, the current SOTA LLM's don't produce code that I'd be satisfied with. I kind of feel like LLM's are really really good at hacking on a messy and fragile structure, because they can "keep track many things in their head"

BUT

An LLM can write a PNG decoder that works in whatever language I choose in one or a few shots. I can do that too, but it will take me longer than a minute!

(and I might learn something about the png format that might be useful later..)

Also, us engineers can talk about code quality all day, but does this really matter to non-engineers? Maybe objectively it does, but can we convince them that it does?

I'm honestly not getting the human story thing when it comes to music and maybe art in general. I mean I get what it means, but I don't think it describes why people enjoy art.

To me, it seems more like people place their own meaning in art. A particular song might remind one individual of the good times they had in their teens, while the actual meaning of the song is completely different.

Bachs 5th symphony (or whatever) might be extremely annoying to someone because they had to listen to it every day at work.

And what exactly is the meaning of jazz fusion? I really like a good solo, but a lot of people hate it, they need to hear a voice. (though I don't particularly like the signature Suno or Udio solo..)

I found this ai track on Spotify that I unironically enjoyed. I listened to it every day while working on reviving an old passion project, which became its meaning to me. The tune, a long with its album with random disparate suno generations was taken down.

I'm not sure if I have a point here, but something is off with the story thing in art to me from a consumers point of view. Maybe from other artists as consumers point of view?

I built an open to "game engine" entirely in Lua a many years ago, but relying on many third party libraries that I would bind to with FFI.

I thought I'd revive it, but this time with Vulkan and no third-party dependencies (except for Vulkan)

4.5 Sonet, Opus and Gemini 3.5 flash has helped me write image decoders for dds, png jpg, exr, a wayland window implementation, macOS window implementation, etc.

I find that Gemini 3.5 flash is really good at understanding 3d in general while sonnet might be lacking a little.

All these sota models seem to understand my bespoke Lua framework and the right level of abstraction. For example at the low level you have the generated Vulkan bindings, then after that you have objects around Vulkan types, then finally a high level pipeline builder and whatnot which does not mention Vulkan anywhere.

However with a larger C# codebase at work, they really struggle. My theory is that there are too many files and abstractions so that they cannot understand where to begin looking.

The responsiveness of windows 2000 in a vm is insane. It feels like every action happens instantaneously.

Contrast this with the "os" of my LG oled monitor. It seriously takes 5 seconds to open the settings menu.