![Great job 👏 [Twonks]](https://lemmy.world/pictrs/image/69fd8e3b-4aea-4e91-addf-bd8fafc25787.jpeg){kind=link}
Twonks | Bluesky
Transcript
TW😶NKS
A comic in four panels:
Panel 1. White text on black
AI Design Logic
Panel 2. A guy sits in a restaurant at a table with a checkered table cloth. A waiter stands near, hands behind back waiting attentively.
Guy: Get me a cheese pizza
Panel 3. The waiter returns with a pizza in hand.
Panel 4. The guy gestures proudly at the pizza. The waiter looks less than amused.
Guy: Wow, look what I made!
It reminds me of what painters said about photography early on. While they toiled for hundreds of hours to paint a realistic image a dude with a box and zero skill beyond maybe some chemistry could snap an even more realistic image. Photography was not considered art for a very long time by the formal art community.
AI slop is AI slop, but if you compare it to a film director explaining to a concept artist or director of photography what he wants, then selecting and refining examples and concept art and sketching out a storyboard, there is little difference to someone using AI to make some film. Just that the latter will typically be much worse because anyone can do it now, they won’t have studied film making, and AI isn’t very good at it so far. But either can be art, or it can be garbage.
Also with advances in hardware and better tools and software and models, we could see a new type of art form or medium. Something like the holodeck in star trek. Like a movie that is more interactive, like a role playing session in VR. Told to you by a narrator but you can interject and steer or derail the narrative. So the “directors” become more like world builders or campaign designers.
The sheer endless tsunami of this dreck makes it meaningless.
I’m not sure more people having more access means things are now worse. Like more people have more means to create art today than 500 years ago and yeah there’s a lot of uninspired or derivative content but also there’s a lot more legitimate artists making passionate and thought provoking work.
Yeah I imagine what you’re saying will come to pass although I think it’ll look more like parlor walls from Fahrenheit 451 for the majority. I think my main issue is when a human is totally uninvolved in the process. So far it seems pretty noticeable when there’s someone with an actual idea they’re trying to create vs there’s no mind behind it.
Haha love the reference to Fahrenheit 451. There is also a good illustration in the horror story from _9MOTHER9HORSE9EYES9 with people sinking into some kind of experience stream, similar to wireheading. Or in “Systema Delenda Est” about postbiological life there is the concept of “Elysium” virtual worlds where people go in and never want to come out again, because they live in a narrative that is always perfect for them. Instead of slop you have perfection that is irresistible. Star Trek had Barkley becoming holo-addicted. Cheese Pizza is really bad for you too lol.
You could argue that something like computer games the story is ultimately told by the player and his decisions and actions, and the world should just react. There is a fundamental limit to storytelling in computer games, which only tabletop role playing games solve with an intelligent narrator or game master. I do think LLMs can fill that role but not very well. But better than what open world games offer today. Not better than a well written movie.
I think ultimately the will or mind should be unbounded sans the ability to willfully cause harm to other minds. It’s purely due to limitations of technology that we are in our current predicament. The famous prosthetic engineer Hugh Herr made the argument it was not him who was disabled but our technology that was disabled. Now what minds do when they are unbounded by the reality around them is their own responsibility. If you’ve seen the show pantheon then things could turn out similarly aka minds in multiple embodiments exploring physical and simulated realities or otherwise experiencing whatever they desire without limitations beyond the limitation of willfully causing harm to others.
Ideally minds unchained from limitations wouldn’t desire to cause harm anyways, but there’s some real jerks out there.
Your analogy doesn’t work. Anyone could take a photograph, just like anyone could paint something, but making it art, eliciting a human reaction, requires skill.
And by the way, not every moment in history when a technological breakthrough threatened a well established craft, can or should be compared to the current state of AI, if anything because it paints a picture where anyone opposing change is some sort of luddite. “Change” is not always a good thing, and there are more than a handful of inconvenient examples: atomic vs conventional weapons, cigarettes vs herbal treatments for anxiety, leaded vs unleaded gasoline for increased engine performance, glass and paper vs plastic in food containers, etc. For each one, incessant marketing campaigns were launched, touting their advantages and how they would improve the world as we knew it. World Expos were organized around these things. People were legitimately excited.
The truth wasn’t quite that. The US found itself in a tight spot when they lost the monopoly on nuclear weapons, which triggered a histeria across the country that lasted for decades, and even nowadays, “nuclear deterrence” conditions geopolitics to the maximum degree. Cigarettes, turned out, weren’t as good to calm nerves as they were killing people. Leaded gasoline, which was the answer given by refineries to the question “what if we stop improving refining processes and start adding literal poison as an additive”. And microplastics are the latest in all the ways humans have found to screw the food chain.
His analogy works perfectly fine and is actually incredibly apt. Photography did threaten a large number of painters livelihoods.
After the grand scale AI is. But that’s more a matter of how things are just on a grander scale as we have become more connected in more optimized as time has gone on.
Technological improvements tend to be larger than those that preceded them when it comes to replacement and displacement.
And if you’re going to put the bar of art as simply eliciting a human emotion reaction or feeling. Then by your own definition AI generated art is art.
It generates all sorts of emotions, reactions, feelings, Thoughts, Arguments, considerations and philosophizing all across the world.
Ai art has probably been one of the largest points in the last couple hundred years in which this many people have genuinely stopped and considered. What is art.
I’m not a big fan of AI art at all, but I’m also not a big fan of trying to discredit something just because you don’t emotionally agree with it. I would wholeheartedly agree that 99.99% of all AI art is low effort slop and absolute trash. But I wouldn’t discredit anyone who actually attempts to do something legitimate with it.
Functionally the entire argument against AI art being art is the same one against photography or modern art not being art.
No, it did not. The story is missing a critical component: anyone who had the money to commission a paint would still do it. Anyone who didn’t, wouldn’t. If anything, photography expanded the portrait market to those people who couldn’t afford a painting.
I didn’t say that. I said that art requires skill as well, and skill in art, in order to elicit a human emotion necessarily involves a purpose that is worthwhile. Otherwise everything, from trolling to taking a dump, could be considered art, and it is not.
You could argue that prompting a LLM is a skill and it could carry a purpose, but I fundamentally disagree with that and I haven’t seen any AI creation that would lead me to change my mind just yet. Let’s put it this way: if I commission a paint, describe in the highest detail what I want in it, the painter does exactly what I asked them to do, and it turns out great, I wouldn’t be an artist, I would be a patron. Now, if the painter decides to just copy and rehash bits and pieces from other paintings with no purpose other than to hide their own inability to create, they would be a plagiarist, not an artist.
Going back to photography, it honestly sounds like your baseline is the skill required to capture something with the highest fidelity possible. Then yeah, painters would be thoroughly fucked. But photography as an art is way more than that, it’s about capturing a moment in the photographer’s own terms. And in that regard, it doesn’t differ from painting at all, unless you wanted to argue that Manet or Dali were trying to portrait the world as everyone sees it.
But AI art? Come on man, most people can barely verbalize what they want for dinner, how are they going to prompt a LLM to create art? “Paint the unbearable lightness of being, make no mistakes”?
Not every art piece is even physical or tangible in any sense. The art is in the conception or the idea. There is indeed a skill in constructing models as well as using them but ignoring all of that there’s still the artistic touch in creating thought provoking or impassioned concepts irrespective of the tangible form they’ve been given. Although one can most definitely use a medium to impart meaning.
It was never my argument that people who have issues with new technology are necessarily luddites or that AI or really any technology whatsoever is absolutely problem free. Everything has its pros and cons, for example nuclear gave us access to functionally limitless clean (compared to coal and gas) energy as well as more powerful weapons. One unavoidable rule or law of reality is we don’t get something for nothing.
Ultimately what is and is not art was, is, and will be an open question each individual has to answer for themselves. Even today an individual is free to view photography not as a legitimate art form or digital art or CGI and so on.
We weren’t talking about constructing models. And where’s the skill in “using” them in this context, when the user is eventually given the exact end product they asked for, and it is as original as rehashing the training set can be?
I see that you went with the nuclear example, because all of the others seem straight up indefensible in comparison.
Problem is, I wasn’t talking about nuclear energy. Even if I did, development of nuclear energy didn’t require creating nuclear weapons at all. Yet they did, and it’s a great example of how technological achievements aren’t always a good thing.
I mean if two painters make similar enough strokes with similar enough paint you get two essentially identical paintings. Hunting the solution space for an NN can be like working through all orderings of a card deck, that’s to say very far from as trivial as you’re making it sound and about as silly as the example I gave above.
I’m not sure exploring just a few examples in the ocean of essentially infinite examples or analogies does much for us both beyond picking something and arguing which is more accurate. Obviously you’re arguing the more useless examples are more accurate and I’m arguing NN have real world use and vakue as shown at least by alpha fold. I would say your basis of argument is more incorrect from the reality we’re sharing and I’m well aware of the pros and cons of what we’re dealing with.
Neural networks are a valuable technology for humanity, to say otherwise is to be blind to reality. That said there’s plenty of actually legitimate arguments one could wage about their current development and use.
Tldr: you be hating too hard, chill my friend.
Who are you responding to?
I mean I don’t know who you are personally. . . Like what details about yourself do you expect a stranger to know?
You didn’t address a single one of the points I made. I wrote a philosophical argument about technological adoption not resulting in a net benefit for humanity , historically speaking, and you chose to meander about solution spaces in neural networks and how useful these NNs are in protein folding problems? What?!
This interaction is just… weird. Assuming that you actually read my comments, you don’t seem to be able to synthesize ideas from them and formulate related answers. This last comment of yours is just odd and completely disconnected from the thread as a whole. To be honest with you, it sounds like a bot wrote it.
I literally brought up alpha fold for that exact point. Anyways, I can see you are anti-neural networks no matter what from your history. Are there any other families of algorithms you are prejudiced towards or is that the one and only?
You forget that early photographers were almost exclusively painters.
Funny eh.
Arguably more so chemists but yeah I’m aware. Most / all of their paints were made by hand themselves as well, part of the reason it was a sensible transition to film production and manipulation.
Well painters at that time were definitely chemists.
That is a good argument.
It’s more so historical fact. It was denounced for like nearly a century if not longer depending how you’re measuring. It’s similar to AI in that it decreases the mechanical effort one needs to create content but also itself imposes new required skills to develop that content. There’s definitely a hierarchy in AI content and so far at least to myself it’s very clear when there is a person actively trying to create something vs a completely automated content pipeline. I would challenge anyone who thinks all AI creations are nothing but slop to create something themselves equal to or better than some of the better works I’ve seen. Similar to a challenge of a traditional painter to create a compelling photographic composition.
Maybe to add to this I think it’s worth noting the creative demand of designing and training a model. It’s not without inspired thought. I’m curious to see the code as well. Amongst people who program there can most definitely be beautiful and ugly code in the same way there can be a pleasing or unpleasant movie scene or painting.