Why is it strange? To me, it could only be strange if one was considering that someone planned this, and then yes, it's hard to come up with anyone who would benefit from this. OpenAI looks dumb, but also their models sound impressive, Chinese models look dangerous, but also useful. No clear winner.
Thing is, I don't think anyone planned this, so to me the timing isn't strange at all. The models really were getting close to being able to have a big cybersecurity impact (I started seeing that after teams were reporting their Mythos usage), and an event like this is not so surprising, given that.
I like Anthropic, I don't think all their talk of safety is bluff and bluster, or at least, I want to believe that the people who left OpenAI because it had lost its focus of helping humanity still want that to be their main goal. However, yes, it seems that business fears are once again causing those in charge to turn "we want to help humanity" into "we are the only ones who can help humanity, and therefore we need to be the most profitable, and the only survivors".
If you want the former ideal to survive, at Anthropic and outside of it, you need to be willing to collaborate beyond profit incentives and recouping capex. Show other labs a commitment to research and community and they will follow. Better to bring teams together rather than implicitly say you distrust them, pushing them that way instead.
"What's your unique contribution to the world" is a good question, and I've been thinking about it a lot. One framing I very much like is thinking about the priceable versus unpriceable contributions of a person. As an example, your software engineering ability is priceable (it's your labor). Your value to your parents isn't priceable. The problem is, right now the default assumed value of the latter is 0. Our society only tries to put a value on priceable facets of a person. There's reasons for that, wisdom of the crowds and whatnot, but of course it also means that if an AI can do that thing, economics drives society to adopt AI over employing people.
It's possible to set that latter value to be nonzero. You can't use free markets to set it because they necessarily cannot see what price to set, so you kinda have to guess, but IMO, it isn't zero, and I'd hazard to say it's positive.
Silly nitpick but the free market absolutely can, and has in the past, put a monetary (positive)value on your person.
We have of course abolished slavery since then.
Which is a weird way of saying I don't think we necessarily should put a monetary value to people. Instead we need to realize that money isn't everything in life. So much of life is unpriceable, and that's as it should be. Attempting to price the unpriceable necessarily makes a mockery of the unpriceable thing.
In my follow up pieces in the series, I detail a way to make the economy actually see a lot (not all, but way more than before) of that value. I'm pretty proud of it. It might be politically hard, but it's theoretically very sound.
On your start page, under all posts (which seem to be chronological), it sits below "The room the economy can’t see", so I did not associate it with the follow up.
Ah, thank you! I kind of assumed it would be coming soon as this post was dated as “today”, and it seemed illogical for the next post to already be out. :)
That's true, I think the real problem is a liiiiittle more nuanced than just the great stagnation. But that's the graph people know. The issue can be read out in more detail from other graphs too, I'm working on an updated piece. I mention georgism at the end of the piece as well.
But do you agree with the theoretical framing? People not being able to not work forces them to find something. It's like an iron ore deposit being forced to be mined rather than being mined when it's needed?
I don't really. It smells to me like the lump of labor fallacy. Labor participation results from supply and demand and it going down does not necessarily indicate less demand, but can also indicate less supply. That's what you'd expect as countries and the world become wealthier. Leisure is a normal good.
The difference is that we are asking the economy to create jobs even when it doesn't "want" to. It still will, the lump of labor fallacy is indeed a fallacy, but that doesn't mean this isn't causing a lot of harm.
We aren't draining the earth's oceans (yet) because its supply is more than the earth needs right now. One day we might, but how much we draw is decided by the economy.
But for labor supply is forced, capital flows there when it shouldn't, politics chases it. The economy, ever adaptable, provides jobs where it can.
The ability to take leisure is still conditioned on labor. To take time off, you have to save up. To save up, you have to work. Labor is distorting the real value of leisure if the production of that leisure needs less labor (leisure time still requires you to be fed, housed, and more if you go on holiday, travel, or consume to keep yourself entertained or educated).
That's not a distortion, that's just what the choice is.
When a new gaming console comes out and someone works extra shifts to afford it sooner, that's not a distortion.
I think your mistake, based on how you're talking about the "real value of leisure" is thinking there is a single value of x. That's labor theory of value era thinking. There is only marginal utility and marginal rates of substitution.
Trading leisure for consumption via labor is the choice, not a distortion.
I agree, and in some sense I argue this at the end, but I didn't want to make the piece even longer.
In fact if you look at the economics of this policy over the long term it leaks capital through rents, so it doesn't actually work forever.
A sovereign wealth fund can solve this (essentially an expansion of the "buffer" I wrote about). That functions as almost exactly what you want, giving a stake in every company to every worker (except broader, to every citizen).
If you buy the framing that labor is currently a distortion, you wouldn't want to only give stakes in companies to workers of those companies, if you do that the distortion will continue.
Perhaps, but this is actually a both a somewhat novel framing of the problem and of the solution. If academic economists agree with this theory and subsequently get behind it, it could be like climate change.
Except this change isn't anti growth and is a lot simpler to fix than climate change. It can be bi-partisan, if framed right, we have to at least try.
Thing is, I don't think anyone planned this, so to me the timing isn't strange at all. The models really were getting close to being able to have a big cybersecurity impact (I started seeing that after teams were reporting their Mythos usage), and an event like this is not so surprising, given that.
reply