What do we want computers to do?
lmnt.me
lmnt.me
"Practice any art, music, singing, dancing, acting, drawing, painting, sculpting, poetry, fiction, essays, reportage, no matter how well or badly, not to get money and fame, but to experience becoming, to find out what's inside you, to make your soul grow."
Over-reliance on computers "doing the hard stuff" stops the human act of becoming.
Which is exactly what the personal context thing is about, so what the hell is this complaining about?
As for the parenting advice, it sounds like it comes from a poorly written textbook written by someone who watched an TV advert about a child's bedtime. Note: parent of three / professional terrorist negotiator. That's not how it goes down most of the time for most parents. The reading and activity time is mostly opportunistic until they get to the point they read their own books and want to talk about them. Best to spend time teaching them to read when they are not burned out and exhausted as well.
Much as the article describes, we are using contemporary computing technology to create infrastructure that allows us to live in the illusion of creating. The infrastructure is inherently cynical and consumption oriented, so it produces cynical consumers. A proper computing infrastructure, namely an environmental computing infrastructure, should produce producers.
Users of ChatGPT are serfs not owners. Their use of ChatGPT provides value to ChatGPT the corporation. Today we live in techno-feudalism[2], and not a time of tech ownership. So the serf gets wheat and cattle, the king stays in his castle.
I am against that these tools are actually tools of consumption. I am for software production ownership.
[1]: https://hothardware.com/news/adobe-responds-to-creator-outra... [2]: Yannis Varoufakis, former Greek Finance Minister, coined the term.
A classic desktop model, where the OS is a single application, user-programmable with ease like modern Emacs or classic Smalltalk workstations, of course networked with other systems offering the same kind of interaction, so a kind of "environmental computing" but personal that's pro-human. Because you are not slave of someone else created automation with more or less limited personalization, you are the owner, the commander of anything belonging to you.
The case of music is quite fascinating. YouTuber Rick Beato is following it very closely. He seems to think that quite soon people will listen to AI generated music even knowing it is AI-generated music. They will listen to it because they'll enjoy it and they will not care who or what made it. I personally think he is right. Music can be enjoyed in a way no other art can, in that sense it might become kind of a drug.
Now some people think that will never happen because AI music will always be bad, but recently he posted a video in which he tells us that his kids can immediately tell an AI-generated song from a normal one. He is baffled because he himself can't do that. But he also tells us that one of his kids thinks that he probably won't be able to tell it anymore in six months or something. In other words, a young person with such a good ear is confident that AI-generated music will soon be if not good, at least indistinguishable from human-made music. I think that's indicative of where things are going.
I think they are mistaken, and that this is the wrong way to go thinking about this.
If you're against AI "doing art", it mustn't be because "it's not good". Because art is very subjective -- so much that it's almost impossible to define what it is -- somewhere someone will like AI art. And the procedures and models will get better, too. I can envision a (very near) future where blockbuster movies and hit pop songs are entirely written by AI... it wouldn't be too different from the present, anyway.
No, if you're upset about AI art, it must not be about technical quality, but about human connection. Art is a human activity, by humans, for humans. Even if AI "gets better", I don't want us to be cut out of the loop.
Art is not something to optimize and automate.
(All of what I've just said is debatable, of course. Like art!)
some music can be passively listened to and is built with consumption in mind (e.g., background music in games or restaurants).
other music has a message and (as pretentious as this sounds) is about communication and connection between the artist and listener.
it wasn't obvious to me until I started thinking about AI art, but so much of the value I get from a piece of media is knowing that another human made it -- i can be impressed with their skills or just be glad that somebody felt the same emotions I felt (and was able to express them).
I'm sure AI will create technically amazing pieces of art someday, but it will never be able to attach a relatable backstory to it
I feel like almost none of the value I get out of a piece of media is knowing who/what made it. I think media either stands on its own merit or it doesn't.
It could be made by monkeys on typewriters, AI, whatever. The media itself is what matters to me, or makes me feel emotions, etc. Not the fact that a human made it. It just so happens that I haven't experienced media made by AI (or monkeys on typewriters) that is as good as the stuff made by humans. But I don't think that'll hold forever.
A shitty movie still sucks even if made by someone I am impressed by or relate to. One of my favorite book series as a kid was written by someone I now have zero respect for, yet I can and still do enjoy the books. And I'm sure I'd still enjoy those same books if they were written exactly the same way but by a different author or AI.
However, this is the first time I've really thought much about it, so my opinion might be different after some further introspection.
imagine that there was an AI that could generate infinite funny cat gifs. do you still like the gifs as much knowing that the cats are real? is it any different from people posting staged videos and trying to pass them off as real?
I feel the same way.
Increasingly, it seems we're making it possible for AI to create art (movies, novels, paintings, music) and eventually be involved in everything related to art: reviewing said art, commenting about it online, maybe even having heated discussions about it.
But art is about human connection. It's not something we want to automate.
The article tangentially mentions something about art becoming accessible to non-artists, i.e. people who wouldn't be able to "create" otherwise (what the author calls the "illusion", or "faking it"). But maybe art isn't supposed to be "democratic" in this sense; maybe those who cannot directly create in their preferred art form should instead enjoy it, or discuss it -- or better, practice more and get good at it the hard way?
I think your human bias is showing. I suspect there will be (specifically) 1 AI - that gains that mindshare. It's all branding! There's a gaggle of pop stars, and one taylor swift... likewise there will be heaps of AI but only one pop_star9000(version3.1415.9265.3589.7932.3846.2643.38_custom_sauce11)
But I would listen to a robot if and only if I believed it could tell me of its own experience... By relating an alien experience with metaphor in rhythm to a popping dance groove.
Isn't this almost equivalent to saying "I would listen to a robot if and only if it was human?". A robot that could tell you its experience would be a living thing. An LLM cannot tell you its "own experience" because it has none. You'd be asking it to be human, or human-like (or even animal-like).
Maybe it's not too bad communication to use templates when creating emails intended to be read by humans, but to have those autogenerated emails to say things like "thank you" or "sorry" in any form certainly is.
Yup anyone who reads a pre-written bedtime story to their kids is a bad parent /s
Even before ChatGPT, I was using davinci in the OpenAI playground to have a lot of silly fun with my daughter. Totally dumb stuff like asking it to create a menu for an alien restaurant with farts as an ingredient, but it had us rolling.
So, "What Do We Want Computers to Do?" - no matter what your answer, AI makes that question a lot more interesting and raises the ceiling of possibilities. I just wish the author (and others like them) would stop moralizing.
Not a bad parent, but definitely something off about it.
Or rather, my personal experience: reading bedtime stories to my daughter is a moment of connection. She loves this. I'm sometimes tired, and could use some prompts -- which I already have, the vast mental collection of books I've read and movies I've seen.
I don't want to automate this. Suppose I could have ChatGPT improvise a story within some parameters and read it out loud with my synthetized voice -- seems dystopian to me. I want to be the one reading the stories to my daughter, and making them up.
It's about bonding and connection, not about efficient storytelling!
> Even before ChatGPT, I was using davinci in the OpenAI playground to have a lot of silly fun with my daughter. Totally dumb stuff like asking it to create a menu for an alien restaurant with farts as an ingredient, but it had us rolling.
This to me sounds like a good use of software tools. Some computer time with your daughter, making the computer do fun things. Why not? It's not a replacement for bedtime stories, it's something new. I don't think this is what the article is criticizing.
While "artistic creation" is hard to demarcate, I think fooling around with a fun tool is not it. Asking an LLM or similar tool to create an alien menu with farts as ingredients, just to watch it do something silly, is not art-creation. Most people can do it without the LLM, too.
From the article:
> I’m not against playing with these generative AI tools. For months, I played with Midjourney, and I still occasionally play with ChatGPT. I say “play” because they’re excellent toys.