I agree, but the crux of my post is that it doesn’t have to be that way - it’s not inherent to the training and use of LLMs.

I think your second point is what makes the first point worse - this is happening at an industrial scale, with the only concern being profit. We pay technocrats for the use of their services, and they use that money to train more models without a care for the deviation it causes.

I think a lot of the harm caused by model training can be forgiven if the models were used for the betterment of quality of life of the masses, but they’re not, they’re mainly used to enrich technocrats and business owners at any expense.

You can lose respect for me if you want; I generally hate LLMs, but as a D&D DM I use them to generate pictures I can hand out to my players, to set the scene. I’m not a good enough artist and I don’t have the time to become good enough just for this purpose, nor rich enough to commission an artist for a work with a 24h turnaround time lol.

I’m generally ok with people using LLMs to make their lives easier, because why not?

I’m not ok with corporations using LLMs that have stolen the work of others, to reduce their payroll or remove the fun/creative parts of jobs, just so some investors get bigger dividends or execs get bigger bonuses

I’m generally ok with the concept of externalizing memory. You don’t need to memorize something if you memorize where to get the info.

But you still need to learn how to use the data you look up, and determine if it’s accurate and suitable for your needs. Chat gpt rarely is, and people’s blind faith in it is frightening

It’s changed by job: I now have to develop stupid AI products.

It has changed my life: I now have to listen to stupid AI bros.

My outlook: it’s for the worst; if the LLM suppliers can make good on the promises they make to their business customers, we’re fucked. And if they can’t then this was all a huge waste of time and energy.

Alternative outlook: if this was a tool given to the people to help their lives, then that’d be cool and even forgive some of the terrible parts of how the models were trained. But that’s not how it’s happening.

You almost got the joke. You were right there

but the world is more complex than that

Proceeds to give the absolute most naive surface level interpretation

Ok then.

It’s worth mentioning two things here.

I’m not inherently against LLMs and AI.
Im against putting the power of LLMs in the hands of the employers instead of the workers. People self-hosting their own free LLMs to make their job and home life easier? I’m all about that, and I can even forgive the theft and energy usage to an extent.

And also that I’m a developer in this space - I don’t train models or sell them directly, but I make products that use LLMs to increase productivity. I know I’m part of the problem, but I was transfere onto the project and my job is simply too good to quit over it, so I’m a hypocrite to some extent. What people in this space are trying to do is absolutely replace workers so that businesses can save on payroll and increase margins. They don’t say it, but it’s telling how they dance around the topic.

What I want is regulation that protects workers more than employers

No, but it follows. It’s what the open market demands.

I think that the market (profit) incentives the mining, and that lax regulations simply fail to curb that incentive. We can see from the war on drugs that simply regulating or criminalizing something won’t stop it if there is a market, it just drives up prices.

But in any case, this aspect is a lot less ambiguous for LLMs because selling the use of an LLM is legal, and the sellers use the money to train new LLMs.

So let’s create regulations around LLMs and enforce those regulations, before it starts to really affect the job market and environment.

In both your examples, you seem to assume that the harm is already done and that there is no continued harm.

But in both cases the harm isn’t finished; the blood diamond mine owners use the continued sale of blood diamonds to fund their continued mining operations, and LLM providers use the sale of LLM use to fund the continued training of new LLM models.

Regardless of if you think that buying second hand blood diamonds increases overall demand in the market (which blood diamonds sellers benefit from); it is clearly the case that selling (and reselling) LLM services benefit the LLM providers, and we can trivially see that they’re training new models and not making amends.

I largely agree with you, and I definitely appreciate that you’re being very civil in our discussion.

There are a few points I’d like to clarify and maybe counter:

I think you’re putting too much burden of responsibility on the workers for decisions that the employer makes. To just say it’s a liquidity problem is ignoring how many people have a liquidity problem and the sources of that problem, and the responsibility that employers should have to their community. I agree that some degree of turnover is ok, but I don’t think that’s what we’re talking about.

I agree that the entire [field] won’t just vanish, but I believe that the increase in productivity means that they’ll need way fewer workers.
this isn’t just affecting fine arts and support:
This is also affecting things like technical writers, marketing, copywriting, programming, paralegal, even diagnostic medicine. Pretty much any office job is in the line of fire.
And when the spread is that wide, even 10-15% of the workforce is devastating to industries and even the economy as a whole.
And the spread is only gonna get wider as they introduce “agents” who are capable of making “decisions” autonomously, so you don’t need a human to tell the AI what to do, and then do something with the output.
Yes the Luddites never mechanized, that’s the thing they were fighting against. They couldn’t all move to complex textiles, because the market wasn’t there for it, if they lowered their prices enough to generate the demand then they couldn’t recoup their time and material costs.

Wrt supply/demand, an increase in supply drives an increase in demand through a lowering of prices. This is the foundation of microeconomics. It doesn’t really translate to the messiness of IRL, but it’s still close enough that it shows that bad things will happen.

In the end it comes down to what the LLM producers are promising. They’re promising to be able to do all this. Idk if they can actually fulfill their promises, but I think it’s crazy to wait and see if they can before moving to prevent it. They’re saying they’re gonna do it, let’s make them not

Any artists that do lose their job are probably mostly ok with it anyway, since it’s most likely going to be graphical drivel anyway.

Replace “artist” and “graphical”, and you just described most jobs. I don’t think most people are ok losing their jobs even if those jobs aren’t especially inherently rewarding; they’re getting paid for their area of training. They’re not just gonna be able to find a new job because in this hypothetical, the demand for it (a living human doing the work) is gone.

I consider this an increase in supply because it’s an increase in the potential supply. Productivity increases (which is what this is) mean you can make more, which drives down the price, which means that artists get paid less (or just get replaced).

Remember: if you 10x the productivity of an employee, that typically doesn’t mean you produce 10x the product, it typically means you need 1/10th the employees. That payroll saving goes right into the pockets of execs.

Also wrt luddites, they weren’t wrong. It did absolutely demolish their industry and devastate the workers. It’s just that the textile industry was only a small part of the economy, and there were other industries who could absorb the displaced workers after they got retrained.
LLMs threaten almost every industry, so there is a greater worker impact and fewer places for displaced workers to go. Also now workers are responsible for bearing the costs of their own retraining, unlike back in the day of the luddites.

Setting aside “there is no ethical consumption under capitalism”, which is a debatable for another time:

I don’t totally agree with your assessment of 2nd hand sales: it’s not ethical positive, at best it’s ethically neutral, because demand trickles up the market. I could go into this more, but ultimately it it’s irrelevant:

The 2nd hand LLM market doesn’t work like that because LLMs are sold as a service. The LLM producers take a cut from all LLM resellers.

You could make a case that self hosting a free open source LLM like OLlama is ok, but that’s not how most LLMs are distributed.

But buying them funds the diamond miners to mine more. That’s the point.

By participating in the market, you’re perpetuating blood diamonds.

You’re insufferable.

I know the analogy isn’t a perfect fit for LLMs in general. Analogies never come close to describing the entire thing they’re analogs of, they don’t need to.

It doesn’t matter because this is a suitable analogy for the argument. This is how analogies work.

The argument is that because the harm has already been done, it’s fine to use LLMs.
That same argument can be made for blood diamonds, and it’s untrue for exactly the same reason:
Because buying the use of LLMs (which is mostly how they’re used, even if you pay in data instead of money) is funding the continued harmful actions of the producer.

I can’t believe I have to explain how analogies are used to a grown ass adult.

Ok cool, but this is an analogy. Why are you defending the use of AI by megacorps by objecting to irrelevant parts of an analogy on technicality?

Even if the diamond mine owners stop mining, it’s unethical to buy their stockpile of blood diamonds.

Also, there is a cost besides electricity - the theft of artist’s work is inherent to the use of the model, not just in the training. The artist is not being compensated whenever an AI generates art in their style, and they may in fact lose their job or have their compensation reduced due to artificial supply.

Finally, this is an analogy, it’s not perfect. Picking apart incidental parts of the analogy doesn’t really prove anything. Use an analogy to explain a problem, but don’t pick apart an analogy as though you’re picking apart the problem.

It’s not a perfect analogy, models ape the work of artists and take their jobs; it’s like if the diamond was bloody, and as long as it existed, the miner’s family not only didn’t get compensated for the loss but we’re also prevented from getting jobs themselves.

We’re not righting the wrong, were making the wrongs even worse. At some point you have to just burn the whole thing down.

Use a diamond doesn’t “use up” that diamond.

And yet, it’s still unethical to buy already mined blood diamonds from people who continue to mine more blood diamonds. Funny thing about that, huh