I'm not interested in the question of whether an image I make with AI is art or not. I've written decent fiction in my time. Maybe here and there I inhabit the artistic category of "an unpublished genre fiction author." But I can't draw or paint or any of that stuff better than a reasonably talented 11 year old, and I'm 52. If I had artistic pretensions it's not happening.

But once in a while I want to see an image that I can't make and, well, AI boom boom yes here we go.

I don't claim that's art. I don't claim that the AI is an artist any more than I claim that a submarine is swimming. Those categories are irrelevant to my interest, which is simple.

I wish I was artistically talented so I could do stuff and now I can do some of that stuff without being artistically talented.

I feel good about this. It makes me feel better in the world.

Are we going to take this away from billions of people to protect people who can draw and have some training?

Hannah Höch would like a word.

It's a wonderful thing to find meaning in one's work, and for the things in which one finds meaning to be well-compensated. But it is no birthright. Contrary to Nick Cave's view, I can absolutely assure you that non-artists in HR departments and nursing stations and factory floors and classrooms often live full happy human inner lives. Those lives are of their own making and do not derive from the artiste class's output.

Manual production of high-quality clothes, tables, and glassware used to be the norm. Generations of people found meaning in these crafts before the industrial revolution changed the economics. People still do these things, only in rare cases as their primary way of making a living. Most art does not sustain developed world middle class existence. Most art is hobby. And that's okay.

The creation of software and AI systems is itself a form of craft-work and soul-work, which many engineers and scientists relate to the same way that Nick Cave relates to music. It is unclear to me why Nick Cave's striving is more important than the striving of engineers and scientists, or why his feeling of what humanity is, is more important than theirs.

Does a picture of a humming bird lack insight? Does collage art lack insight? Do remixes lack insight? Does mass-produced formulaic pop music lack insight?

Maybe. Or maybe some artists enjoy those creative processes and some audiences enjoy the output. Maybe oil painters who critique photography, and photographers who critique collages, and musicians who critique mash-ups, and DJs who critique modern production studios, and, yes, artists who critique the use of AI models in creative processes, are all just being pretentious assholes.

(It is possible I am simply misunderstand Cave. I take most of his writing to be artistic prose. It's possible that these are sincere metaphysics and that Nick Cave does literally believe in some sort of ur-religious "essential and unconscious human spirit underpinning our existence". In which case I think he's got a nutty religion and consider the fact that AI is an existential threat to that religion mostly a net good for humanity.)

Actually, I agree. I think Nick Cave is right about this. I do think this sort of alienation has a cost.

But that doesn't mean that there is any remotely moral case for undoing the green revolution and allowing billions to starve. And it does not mean that the machines which feed those billions of people who might otherwise starve are somehow the root cause of a decline of humanity. In fact, quite the opposite.

And this is the paradox: our alienation from agricultural work is precisely what enables our very existence.

My main observation is that there is a way out of this paradox. As it turns out, you can go out grow some food in a garden, or write a song, or paint a picture, even if that work is commodified and there is no paycheck. The commodification and automation of those industries does not prevent one from engaging in them as soul-work.

The teacher who plays in a band in his garage is no different -- from a "soul of humanity" perspective -- than Nick Cave. But Nick Cave's implicit argument argument demands that he is different, and not from an economic perspective, but from a very soul of humanity perspective. It's extraordinarily off-putting to me in that sense.

Of course, engaging in art as hobby instead of for pay does require free time and a share of returns on our societal bargain. On that note: elites like Nick Cave should be spearheading serious conversations about political economics and labor economics, instead of lamenting the loss of their extraordinarily unique status.

I have to admit not quite sure what you mean, and I do admit full guilt in starting us down the path of "mixed analogies" :). I'll try my best, though.

> Or past a certain point, the results of discovering improved health and ecological outcomes will become inextricably linked with discovering new technologies that cause conflict? What actually drives such a process?

I do think with respect to life-sustaining things -- medicine, pharma, food, shelter, water, energy -- that a combination of specialization and automation is necessary to increase the collective standard of living, and that labor alienation stems from a combination of specialization and automation.

Where I struggle is coming up with an affirmative argument that an artist should benefit from automation of medicine or farming, but that an alienated lab tech or food factory worker should not benefit from automated art.

Another way to look at this is: the less you pay for art-as-entertainment, the more resources you have to buy free time to produce your soul-work (whatever that may mean to you).

> Another way to look at this is: the less you pay for art-as-entertainment, the more resources you have to buy free time to produce your soul-work (whatever that may mean to you).

Ah, yes. The alienated workers of the world will warm their weary souls at the hearth of derivative algorithmic creativity units. The reduced price and efficient delivery of each drone's creativity units will obviously give them more free time.

Perhaps we can even come up with a pill that'll let the drones feel entertained without any content at all. If the side effects are well-tolerated, they can take it before work.

If I were the author I'd never do this because I would want an exit and this strategy + twitter thread wildly complicates any potential exit.

Even though I don't think there is anything particularly morally problematic here (I'm an information freedom maximalist).

The NYT case could take years. In the meantime OAI could choose to go after ToS violators.

The legal system can accommodate more than one unresolved court case at a time. We don't like put a semaphore on related cases or anything like that. (Or, sometimes we do, but guess who you need to hire for many many billable hours to make that happen in your case?).

So, the legal system can accommodate the NYT case against OAI and an OAI case against the author. The operative question is: can the author's pocketbook also accommodate?

(Or, more to the point, can the author accommodate losing access to gpt4? What happens when he wants to launch a new feature or pivot to a new product?)

If you never want an exit then probably doesn't matter.

Market cap of NYTimes is 8B, OpenAI is 80B, MSFT is 2.7T, you do the math.

2024 will be the year of synthetic data. 2025 will be the year of "you know you can use your own brain and type out 100 datapoints faster and cheaper than generating and filtering assloads of synthetic data, right?"

Maybe we can even skip 2024 :)

> We were initially skeptical whether we would get to 10,000 results. But with nightly leaderboard gamification, we managed to break 15,000 results within a week. Out of fear of eating into our productivity, we closed the contest.

I've hosted a few of these corporate data labeling events. If sufficiently gamified / there's a good enough UX, they can be surprisingly engaging. It helps a lot if you have a large employee base though. Distributing results over 5000 employees is exponentially easier than even 50 - in practicality, even larger than the orders of magnitude.

People concerned with data quality from LLMs should really see the inconsistencies we came up with!

> (e) use Output (as defined below) to develop any artificial intelligence models that compete with our products and services. However, you can use Output to (i) develop artificial intelligence models primarily intended to categorize, classify, or organize data (e.g., embeddings or classifiers), as long as such models are not distributed or made commercially available to third parties and (ii) fine tune models provided as part of our Services

I guess if its behind an API and no one discloses the training data, OpenAI can't prove anything? Even obvious GPTisms could ostensibly be from internet data.

I have two issues with those terms:

1. I think that eventually US courts will determine one of two things: that OpenAI et al are guilty of massive infringement, or that these sorts of restrictive terms aren't enforceable. The need that these companies are trying to treat with terms on output seems unlikely to work out in the end. But we'll see.

2. Even if the terms are enforcable, the human review step in the tweet seems like it's make OpenAI's threading-the-needle position here even more fucking difficult to be taken seriously by any jury or judge.

However, enforcing the terms seems real damn hard in the case of small businesses... as long as you're not stupid enough to admit to violating them in a twitter thread, of course.

I think the author is probably safe from legal action for now because I don't think OpenAI is particularly eager to test the enforcability of their terms. And even if they are, doing so in this case is super high risk and super low reward. Still, I wouldn't test it by openly admitting to ToS violation like this. At the very least seems like a good way to get cut off from OAI APIs.

[1] https://openai.com/policies/business-terms

Of course, you can simply ignore it, just like OpenAI is happy to ignore the terms of services on scraped websites and pirated ebooks and so on.

What are they going to do - claim your model is a derivative work of the training data?

> (e) use Output (as defined below) to develop any artificial intelligence models that compete with our products and services. However, you can use Output to (i) develop artificial intelligence models primarily intended to categorize, classify, or organize data (e.g., embeddings or classifiers), as long as such models are not distributed or made commercially available to third parties and (ii) fine tune models provided as part of our Services;

Depending on what kind of model they trained, they might be breaking these terms.

1. I'm with McKenzie on the coworkers aspect of the dialog. More separation from coworkers is better. In the Good Times (2013-2019,2021) it seems "right" to trade some comp for familiarity and good work vibes, and almost... inhuman... not to. But in the Bad Times you're reminded that an Excel formula could cost you not just your job but also a big chunk of your personal social network. Diversification is good.

2. I now realize what both sides of this are getting at is basically: "how to progress from Junior/Mid Engineer to something after that". There are many paths. The conclusion of the article is good, in that respect. Also: you can just stay a Mid/Senior Engineer. That's okay.

3. Call yourself whatever you want/need to stay employable. Be a good colleague/person. Work is work.

Only a very small subset of industry cares about academic publishing, and even within that subset it's only a fraction of groups at a fraction of corps that consider publishing a primary or even secondary objective.

The groups that do care about those things can be good gigs, but are generally not the place in the company you want to be anyways, unless you can get in and out (for good) in <10 years. If you can do something that actually impacts the business -- that is actually useful to other humans -- no one gives a shit about h-indices or kaggle scores. And you'll be better compensated anyways.

> can do something that actually impacts the business -- that is actually useful to other humans

Do not confuse these two. That's incredibly naive.

Read more pure math papers, then you will see the difference. :-)

This is no contradiction: there exist quite pure math papers whose content is very relevant for the mathematics behand ML algorithms. :-)

I do have the impression that the kind of research in ML that is not strongly associated with the recent "machine-learning industrial complex" by now tends to become published in another subject area.

The problem is that AI is weird not because of academia. In fact, right not it has been captured by industry and it is why we've severely slowed down in progress[0]. Most people in the space now are working in industry labs. Frankly, you can do more, you get paid A LOT more (2-3x) and you have less bureaucratic bullshit. But I think you're keenly aware of this industry capture as you're mentioning aspects of it.

I don't want there to be any confusion: I think it is good that industry and academia work together. There's lots of benefits. But we also need to recognize that these two typically have very different goals, work at different TRLs, and have have very different expectations on the time where the work will be seen as impactful. Traditionally, academia has generally been the dominating player in the high risk high reward/low level research space (yes, much more goes on too, but of people that do this type of research, you think academia) while industry research typically is focused on higher TRL because they're focused on selling things in the near future. There's just a danger when you work too closely to industry: you can't have any wizards if you don't have any noobs.

But I'm not sure it is just ML that's been going this way. There's a lot of sentiment on this website where people dismiss research papers (outside ML) that show up here due to them not being viable products. I mean... yeah... they're research. We can agree that the value is oversold, but often that's by the publisher (read university) and not the paper (not sure if I can say the same for ML). But it's a kinda environmental problem because if everything has to be a product you can't be honest about what you did and if discussing the limits and where you need to still improve upon to actually get an product down the line gets you rejected, well... you just don't talk about that.

This is all the "RL hacking" or better known as Goodhart's Law. I've been saying we're living in Goodhart's Hell because it seems, especially in the last 5-10 years, we've recognized that a lot of metric hacking is going on and decided that the best course of action is not to resolve the issues, but lean into it. We've seen the house of cards that this has created. Crypto is a good example. Shame is if we kill AI because there is a lot of real value there. But if you're a chocolate factory and promise people that eating your chocolate will give them superpowers, it doesn't matter how life changingly delicious that chocolate is, people will be upset and feel cheated. Problem is, the whole chocolate industry is doing this right now and we're not Willy fucking Wonka.

[0] More progress looks like it is being made and there is a lot of progress that should have been made but wasn't but these types of nuances are a bit harder to discuss without intimate knowledge of the field. I'll say that diffusion should have happened much sooner but industry capture had everyone looking at GANs. Anything not, got extra scrutiny and became easy to reject due to not having state of the art results (are we doing research or are we building products?)