Hacker News new | ask | show | jobs
by io84 373 days ago
Just like with food: there will be a market value in content that is entirely “organic” (or in some languages “biological”). I.e. written, drawn, composed, edited, and curated by humans.

Just like with food: defining the boundaries of what’s allowed will be a nightmare, it will be impossible to prove content is organic, certifying it will be based entirely on networks of trust, it will be utterly contaminated by the thing it professes to be clean of, and it may even be demonstrably worse while still commanding a higher price point.

3 comments

The entire world operates on trust of some form. Often people are acting in good faith. But regulation matters too.

If you don't go after offenders then you create a lemon markets. Most customers/people can't tell, so they operate on what they can. That doesn't mean they don't want the other things, it means they can't signal what they want. It is about available information, that's what causes lemon markets, information asymmetry.

It's also just a good thing to remember since we're in tech and most people aren't tech literate. Makes it hard to determine what "our customers" want

> If you don't go after offenders then you create a lemon markets.

Btw, private markets are perfectly capable of handling 'markets for lemons'. There might be good excuses for introducing regulation, but markets for lemons ain't.

As a little thought exercise, you can take two minutes and come up with some ways businesses can 'fix' markets for lemons and make a profit in the meantime. How many can you find? How many can you find already implemented somewhere?

Well throw us bone! Can you cite robust examples where private markets deal with this gracefully? Because I can't.

An informational asymmetry that is beneficial to the businesses will heavily incentivise the businesses to maintain status quo. It's clear that they will actively fight against empowering the consumer.

The consumer has little to no power to force a change outside of regulation, since individually each consumer has asymptotically zero ability to influence the market. They want the goods, but they have no ability to make an informed decision. They can't go anywhere else. What mechanism would force this market to self correct?

Businesses with a reputation for honest dealing and good quality attract repeat business.

Why are you so pessimistic that customers can't go anywhere else?

The classic market for lemons example is about used cars. People can just not buy used cars, eg by buying only new cars. But a dealer with a reputation for honesty can still sell used cars, even if the customer will only learn whether there's a lemon later.

Another solution is to use insurance, or third party inspectors.

  > People can just not buy used cars, eg by buying only new cars.
Listen to yourself here. Your solution is "be rich"

So what happens is you either create a cliff or you pull everything down too. Lemon markets for the poor or lemon markets for everyone. Neither is good

> Listen to yourself here. Your solution is "be rich"

I never bought a used car, or any car at all. And I did that before I was rich.

In fact, poor people generally can't afford cars in the first place.

> So what happens is you either create a cliff or you pull everything down too. Lemon markets for the poor or lemon markets for everyone. Neither is good

Huh, what? Reputation still works, even for poor customers. And so do warranties and insurances.

And you seem to imply that regulation can magically make the problem go away? It can't. Typically, regulation in this case raises the prices for everyone by demanding certain features, whether users want them or not. (But details depend on exactly what regulation you propose.)

  > As a little thought exercise, you can take two minutes and come up with some ways businesses can 'fix' markets for lemons and make a profit in the meantime. How many can you find? How many can you find already implemented somewhere?
 
This sounds exactly like what causes lemon markets in the first place. Subtle things matter and if you don't pay attention to them (or outright reject them) then that ends up with the lemon market situation.

Btw, lemon markets aren't actually good for anyone. They are suboptimal for businesses too. They still make money but they make less money than they would were it a market of peaches.

Let me give you an example: reputation can solve the 'market for lemons'.

If you build a reputation for honest dealing and high quality, then people can trust that you don't sell them lemons (ie bad used cars in the original example). This reputation is valuable, so (most) companies will try to protect it.

And that's exactly what's happening with some used car dealers.

You know, the lemon market also applies to new cars, right? And it still happens with name brands. Go read the paper, it accounts for those things.

I mean it got a Nobel prize. You really think they didn't think things through?

The paper explicitly has to exclude reputation as a mechanism.

They did think this through, but that doesn't mean reputation doesn't work in the real world.

I do wonder what would be an acceptable level of guarantee to trigger a “human written” bit.

I actually think a video of someone typing the content, along with the screen the content is appearing on, would be an acceptably high bar at this present moment. I don’t think it would be hard to fake, but I think it would very rarely be worth the cost of faking it.

I think this bar would be good for about 60 days, before someone trains a model that generates authentication videos for incredibly cheap and sells access to it.

Pen on paper, written without consulting any digital display. Just like exams used to be, before the pandemic.

Of course, the output will be no more valuable to the society at large than what a random student writes in their final exam.

Interesting...thinking this through: For text and ideas the information size is often small enough to fit in human memory, and thus containing this is already unsolvable! I can ask the LLM to compose the text of a pitch and then film myself writing it out. Nothing you can do will prove the provenance of those bits was not from the AI.

So I think the premium product becomes in-person interaction, where the buyer is present for the genesis of the content (e.g. in dialogue).

Image/video/music might have more scalable forms of organic "product". E.g. a high-trust chain of custody from recording device to screen.

Fully in agreement with you. There'll be ultimately two groups of consumers of "organic" content:

1. Those who just want to tick a checkbox will buy mass produced "organic" content. AI slop that had some woefully underpaid intern in a sweatshop add a bit of human touch.

2. People who don't care about virtue signalling but genuinely want good quality will use their network of trust to find and stick to specific creators. E.g. I'd go to the local farmer I trust and buy seasonal produce from them. I can have a friendly chat with them while shopping, they give me honest opinions on what to buy (e.g. this year was great for strawberries!). The stuff they sell on the farm does not have to go through the arcane processes and certifications to be labelled organic, but I've known the farmer for years, I know that they make an effort to minimize pesticide use, they treat their animals with care and respect and the stuff they sell on the farm is as fresh as it can be, and they don't get all their profits scalped by middlemen and huge grocery chains.

You're capturing nicely how the relationship with the farmer is an essential part of the "product" you buy when you buy high-end organic. I think that will continue to be true in culture/info markets.