After watching his speech, the first thing that came to my mind is one of the speeches that was given by James T. Kirk, which I had no idea which episode it was in, but I do remember the core meaning of that speech. So I use chatGPT to actually help me to find it by just single one question. And it works! And here's the thing I want to share with the the audiuence.
"They used to say if man could fly, he'd have wings, but he did fly. He discovered he had to. Do you wish that the first Apollo mission hadn't reached the moon, or that we hadn't gone on to Mars and then to the nearest star? That's like saying you wish that you still operated with scalpels and sewed your patients up with catgut like your great-great-great-great grandfather used to. I'm in command. I could order this, but I'm not because Doctor McCoy is right in pointing out the enormous danger potential in any contact with life and intelligence as fantastically advanced as this, but I must point out that the possibilities - the potential for knowledge and advancement - is equally great. Risk! Risk is our business. That's what this starship is all about. That's why we're aboard her. You may dissent without prejudice. Do I hear a negative vote?"
Isn't that pretty ironic, given the speech leads to the spherical intelligences to take him for a ride, manipulate and trick him, and almost kill him and four of his men?
How is that ironic? Saying "risk is our business" doesn't imply you'll be successful. It implies the risk is worth it, and you acknowledge the possibility of negative and dangerous outcomes.
no, because risk and danger is an expected outcome of what he said.
a weak situational irony would be that the crew reluctantly agrees to take on the risks of exploration, but then every single encounter they have from then on is peaceful, contemplative, and non-eventful.
a weak character irony could be that Kirk gets the crew on-board with his gung-ho antics and thirst for the unknown, they all agree -- and then it's discovered that Kirk is actually a huge wimp that was really just putting on a front because he never thought the crew would want to continue after his descriptions of danger.
Leaving aside the bits based on "soul" and "god" which don't move me at all, the rest seems to boil down to something along the lines of: content produced with ChatGPT lacks effort and sense of accomplishment, and somehow threatens the existence of other "pure" content so we have to fight ChatGPT's existence itself.
This makes no sense to me. For example, wide spread and cheap book publishing, and later internet, means that the barrier to entry for writing has gone way down and probably 99% of stuff written is utter garbage: marketing copy, ghostwritten novellas following a cookie-cutter template, all sorts of crap.
Yet no sane person today would argue this was a bad development.
Once again this is the same story as with any technological progress. If we had a time machine and went back to the invention of the wheel I'm sure there would be a bunch of doomsayers trying to stop that too.
You are completely missing the point Cave is making. He is not arguing against ChatGPT as a technology. He is reminding songwriters that writing lyrics is not a meaningless obstruction standing between them and a commodity they are producing. The process of writing lyrics is what makes them songwriters. The attempt to "optimize" this process to such an extent that it removes the songwriter from it is a dead end.
I think we can easily extend the argument to singers. In fact, I think it's more obvious and clear when we do.
Imagine soon, or maybe even already, we have an AI that can mimic any musician's skills with new words. Then, you go to see a singer. But what actually happens is the person on the stage says "here's a song I wrote as it would sound played by Eric Clapton".
Then they press a few buttons, and there's a really good rendition of their words as if Clapton had toiled for several weeks creating new guitar riffs to match them. It's brilliant.
But... you came to hear a singer, see a performance. What you're getting is music, without any of the effort. It might be a good show, but it is not a live music performance.
If this happened, and the event had been sold as a live performance, most people would be rightfully dissapointed.
Songwriting is quite the same, except that it's hidden. We don't see the writing happen, we only hear the result. With a live performance, the (artist's effort)/(audience feedback) loop is immediate. With tasks like songwriting, or painting, or writing novels, the loop is delayed.
The delay matters. People will care a lot less if they don't immediately see that the effort was made by a computer rather than a human. Even if you tell them later, they won't care as much. The human psyche gives huge weight to immediacy. But I would argue there's no difference. If you appreciate the artistry and effort of seeing a human performing live, you should try to extend that to appreciating the effort of an artist who toils for weeks or months behind the scenes before showing their work. Otherwise you're saying that live performance is somehow a more valid form of art.
This is not an argument against AI art or music, by the the way. If the show, the painting, the song, or whatever, is clearly labelled as an AI creation rather than a human artist's, I have no problem with that. However, we are alreading seeing efforts in the film and music industries to remove humans and replace them with AI. It's starting with extras but as the tech improves it will move on to everything else. The big studios would love if they can get away with that. I don't think this is a good thing.
Humans and our tools are, in my opinion, a singular thing.
We evolved _with_ our tools — they afforded us greater access to food, which led to more intelligence, which led to more advanced tools, and on, and on.
A lot of people take this kind of "other" view. Technology is "bad", when really it comes down to how we choose to use it.
It's like Herbie Hancock said when talking about machines and music in the 1980s: an axe can be used to slaughter your neighbor, or it can be used to build a house. They're tools, and ultimately it comes down to what we choose to do with them.
A pretty important distinction between machines in the 1980s and ChatGPT is the amount of human effort that the output requires. It's unlikely that Herbie Hancock considers people who press "start" on a Casio synthesizer to play a factory programmed demo track real musicians.
If someone incorporates a factory preset into a song - sure, nothing's wrong with that. That's when a machine is used as a tool, not as a substitute for human creativity.
> It's unlikely that Herbie Hancock considers people who press "start" on a Casio synthesizer to play a factory programmed demo track real musicians.
Someone at the factory wrote and programmed that song, so he probably would consider it art.
> If someone incorporates a factory preset into a song - sure, nothing's wrong with that. That's when a machine is used as a tool, not as a substitute for human creativity.
It's all just different levels of the devil's company. What if I use an AI chord generator and decide which chords to use in my song based on what the AI generates? What if I have the AI generate 10,000 songs in totality and I choose the best one? What if I don't use AI but compose a song by mathematically computing all the optimal elements of a pop song based on every song that's come before it? What if I hire the world's best violinist and record 4 minutes and 28 seconds of the violinist not playing, could that not be art?
How we chose to use it and what happens to those in the precariat.
The precariat - meaning here those who get displaced from work when innovation happens.
Why is it reasonable that some people get socially disadvantaged when an innovation happens that makes their work much more efficient? Not saying that that always happens but this disadvantage itself, and not the innovation, I would say is a reasonable reason for being a luddite.
That’s very true in a deep way. We have smaller guts and jaws because we tamed fire. We domesticated animals and plants to our needs. If we had to eat raw meat and the 50k year old wild variety of vegetables we’d be on the brink of starvation.
These are two sides of a rather naive understanding of technology that is yet surprisingly common in "tech".
Technology is neither good, nor bad, nor neutral.
Its value is relative. Technology is a means to an end. If you find the end good, and the technology helps towards that, it's good for you. If you find the end bad, and the technology helps towards that, it's bad for you.
Some amount of additional complexity comes from unpredictable consequences of technologies — while someone develops a technology to reach a certain end they wish for, they might inadvertently lead the world towards a different end they don't want as much or even are actively against.
But technology is just a modality of change, a way change happens. And change is neither "good" nor "bad." Some people in technology endorse a "progress" narrative according to which technology is good. It just means that it goes towards an end they like.
That something can be used to do both good and bad doesn't make it "neutral", either. That would mean it has no valence (and probably no influence). Instead, it's both good and bad at the same time for different groups of people.
The atomic bomb, which can arguably be used for "good" and "bad" isn't "neutral." It creates strong path dependence. Now humanity is in a timeline where atomic bombs exist, which is different from ones where they don't. And we can't ever go back to a different timeline (barring some weird time travel shenanigans).
Is it a better timeline? Maybe, maybe not. It seems like it's a more interesting one, in the same sense that people wish others "may you live in interesting times."
I agree with Cave's perhaps overly dramatized response. But overall it's a response to a very early misinterpretation of what statistics-based models are for. Sure, they are capable of producing an output that looks like a product of human creativity, but so are e.g. chance-based techniques (dice rolling to pick a sequence of notes in a scale). LLMs should and will be used as tools, they won't replace human creativity in any meaningful sense.
Doing something challenging has merit in and of itself. I am sympathetic with Cave's concern with the commoditization of works and it is easy to see how someone would have a poor opinion of machine assisted works when inundated with the mediocrity of half arsed efforts created by tools that are still in their infancy. I feel the issue at heart here is that is is easy to create something now that superficially resembles a quality work, but on anything beyond a glance are significantly flawed. People are doing this, for the most part I think they are aware of the flaws and are just exploring the possibilities, but a few get carried away or lack the ability themselves to distinguish between the superficial and the deep.
If a machine could write a book that is better than Lord of the Rings, I would want to read it, and I think it would be dogmatic to dismiss it simply because it was generated by machine. To create such a work though, might require a sufficiently advanced AI that you might be able to meaningfully describe what it did to write it as effort. If it were to go further to a superhuman AI which could knock out books like this with ease on a daily basis, would that diminish those works? Isn't that akin to saying that books from talented authors are less worthy than those from untalented?
AGI might be a modern form of divine 'rest,' taking on the analytical legwork so we can engage more deeply with our creativity. It's mapping our collective subconscious—art, stories, dreams—not as a shortcut, but as a deep-dive tool. This could actually draw us closer together on a soul level, enhancing, not replacing, our collective human journey.
In most arts the constraints yield the technique, where AI is removing constraints faster than most people can even begin to apprehend them. A lot of our meaning came from the constraints we accepted, where we were shaped by the forces we opposed and produced things that were expressions of that experience.
If AI removes these, the effect (without competence) is that there's less to shape us. I see it as a kind of Great Decohering. Maybe we need to build an essential civilizational Arc or time capsule that could survive and be discovered in some future era when some variation of our species develops the technology again so they know there is a there here. Thinking the first major humanoid robot application should be to build moon pyramids.
Tools are tools. ChatGPT is a tool just like a camera is a tool. I'm sure a lot of landscape painters in the late 1800's felt the same as Nick Cave does today. "Cameras take the effort out of rendering images of reality thus robbing humanity of the struggle that in the end equates to love." Tools like ChatGPT further open up the creative process to larger numbers of people. It removes the traditional gatekeepers which is why they are pushing back so hard.
> Tools like ChatGPT further open up the creative process to larger numbers of people. It removes the traditional gatekeepers which is why they are pushing back so hard.
In what way was the "creative process" closed off to those people before?
Not op, but my answer to this would be years of practice and skills. Someone might have an amazing idea, but might have trouble putting their vision on paper.
ChatGPT and generative art helps "idea people" concretize their vision more easily.
The promise of generative AI is exactly this in my opinion; giving the ability to anyone to simply wish an idea into a concrete "thing"
I'm not the person you're replying to, but one can think of a myriad of ways.
Our society is currently structured in a way where the time and money needed to "freely" pursue a creative endeavor is not afforded to most.
I consider myself to be in a particularly "good spot" in my life. But as a new father, I oftentimes cannot afford the spare time between my societal obligations, and fatherhood. This means my creative endeavors take the back seat.
Nick Cave certainly has his heart in the right place, but he is panicking a bit with that letter.
It's certainly possible that creatives in literature, music and image-making can augment their output somewhat with AI and perhaps even more with AGI. This is going to be a problem for them as much as it is for tech workers. In the not-so-long-term it means fewer people will be able to produce "enough" output to drastically reduce the required headcount in places like script writer's boiler rooms, music and graphics production for advertising and in-house corporate productions.
But there's definitely "a feel" to heavy-handed use of AI generated stuff. It's palpable. Much like that "uncanny valley" in robotics. I expect it will remain detectable by people who care far into the foreseeable future. I also expect there's going to to be an expanding long-tail market for authentic works that don't have that "whiff" of AI in them.
Here's what Chat-GPT says:
Nick Cave's concerns reflect a broader debate about the role of AI in creative industries. While AI can augment creative processes, there's a valid point regarding the unique qualities of human-made art. The “feel” or essence that distinguishes human creativity is often nuanced and deeply tied to the human experience. As AI technology advances, the challenge for creatives may not be to compete with AI, but to emphasize these human elements that machines cannot replicate. The market for authentic, human-created art may indeed flourish, as there will likely always be a demand for works that resonate with the human "touch." This dichotomy could create a new appreciation for human artistry, alongside the use of AI as a tool for expanding creative possibilities.
for the past decade or more, the bourgeoisie and bohemian types have cheered and sneered along the lines of "if your job can be taken away by an immigrant, you aren't good at it".
well, now their jobs are going to be taken away by computers who will do it for fractions of a penny, And That's A Good Thing, Here's Why.
I understand the sentiment, but it's the wrong approach. Better to embrace and integrate than to bitterly fight what's inevitably coming. Every technology has had its negative impact, there's always something that dies and can't be replaced. But, something else takes it place and gets brought to life, and that's beautiful in it's own way.
If the technology replaces humans finding their own values and expressing them with a regurgitation of whatever's the most likely to follow a given set of letters in human writing so far...
well, I for one find that repulsive.
I may be nothing but a stochastic parrot, but my life experience doesn't feel like it, and I have less than no interest in consuming output generated by one.
We may all be stochastic parrots, but we are individually so based on each of our experiences. AI however, is still a lot like getting answers from the same person only there's millions of users asking that person. I think once we have better open sourced models and greater ability for non-technical people to modify their AI's behaviour, we will see the innate human creativity shine through AI. My worry is that LLM's will influence how humans write (even while not using the tools), which will affect the models, which will change our writing styles and so on until languages are homogenised and local/personal linguistic quirks are all but gone.
The only way for this technology to replace humans is if people actively choose generative art over humanity's regenerative art.
Do you think you'll be replacing your artistic consumption with mostly auto-generated fare? If not, then is this repulsion you speak of aimed at generative AI, or is it aimed at some hypothetical other people of less discriminating taste who would actively choose to consume it instead of human art?
I guess I'm late to the party, but it is starting to feel as if complaints about generative art are actually complaints about other people.
Damn kids and their rap music. Get off the lawn, right?
I do worry that as generative AI becomes normalized, many people who could have brought a unique perspective to the world and expressed it in their own way will instead be lured down the path of being guided and influenced by those systems instead of doing the work themselves.
Something like the concerns John Taylor Gatto presents in The Seven Lesson Schoolteacher, but with seductive ease of creation as the primary mode of influence, rather than top-down authority.
If it takes them a few decades to work that out, they'll be past the point in life when mastering artistic and expressive skills is most-effectively learned.
It's not hopeless, but it's absolutely a direction I don't want our culture to go in.
From personal experience, it is never too late to master an art ;)
It is my sincere belief that what this will do is raise the bar and accelerate the growth of human potential -- when people realize how mich nicer their world is when aesthetics form a core component, they'll long for something of more substance and in the medium and long-term this will increase the demand for human art.
Additionally, by giving people a taste of what it feels like to be "creative", the supply side will be stimulated as well -- and eventually, with real-time back-and-forth between artist and machine (such as Krea), this will become just another tool in the box of the artist -- leading to a renaissance in the art world.
But, as always - be the change you want to see! Protect our culture by continuing to share your opinions - and your art.
I think you may be thinking of A Day in the Life where "Lennon's lyrics were mainly inspired by contemporary newspaper articles." There was a lot of creativity in that song also though.
>"A Day in the Life" became one of the Beatles' most influential songs, and many now consider it to be the band's greatest work. (Wikipedia)
I can’t disagree with his point about song writing but I think ChatGPT is not going to alter the struggle that we face on a day to day basis. It is just going to amplify our accomplishments. The art that a single artist makes tomorrow might be the product of many lifetimes of human work yesterday.
I remain unstirred. Human creativity is just curation at a smaller scale, I mean to say that even freestyle artists practice and give us only their best stuff. The problem artists have now is that the vague or mysterious origins they once put up as the reason for their inspiration is being exposed as fraud. However powerful their muses were before AI run amok now they find that ambition has turned into dejection, an unwillingness to compete that hard for adulation. You love to see it.
If you really value human creativity you should value when it is used to curate mechanized AI output as well as organic brain mush output. Even pure unfiltered Trump/Biden A.I. debate roasting is only funny for its potential to be further curated and shared, and because that is now easy it means the laughing it produces is less of a laugh? I don't think so.
"They used to say if man could fly, he'd have wings, but he did fly. He discovered he had to. Do you wish that the first Apollo mission hadn't reached the moon, or that we hadn't gone on to Mars and then to the nearest star? That's like saying you wish that you still operated with scalpels and sewed your patients up with catgut like your great-great-great-great grandfather used to. I'm in command. I could order this, but I'm not because Doctor McCoy is right in pointing out the enormous danger potential in any contact with life and intelligence as fantastically advanced as this, but I must point out that the possibilities - the potential for knowledge and advancement - is equally great. Risk! Risk is our business. That's what this starship is all about. That's why we're aboard her. You may dissent without prejudice. Do I hear a negative vote?"
- James T. Kirk Return to Tomorrow (1968)