Unreal is like venture capital or a book advance (or the equivalent in music record deal)
Can you self publish? Sure, of course, have fun.
But if you want the support and infrastructure of a company that understands the business of books, you take a deal and it is just like this: if a bunch of authors get book advances, that is generous to the ones who are unsuccessful, and they can only do that _because they capture the upside of those that are successful_.
Without that, you don't get advances for anyone.
So the point I'm making here: unreal provides variance reduction for all game publishers and yes that disproportionately benefits the ones who make under a million. But they're the ones who need the help!
And in exchange, if you're one of the lucky few, you pay a shockingly reasonable 5% in perpetuity.
If unreal cost money up front, would this have been built? No.
Unreal is saying: hey, we contributed to 1/20th of your success, because you could not have done this without us.
Thus, in the event that you're extremely successful, yes, you'll owe unreal a million dollars. But that's only because you made 20mm and keep 19mm for yourself.
I think Zach Gage (developer of excellent games including Really Bad Chess, Spelltower, etc) says on Adam Conover's podcast that for many people they have difficulty improving at a skill when they have time (or other) pressure
Thus, he always includes a relaxed mode to let someone practice without any stress. Incidentally, he realized that some people only ever play in the relaxed mode!
To be clear, some of the token expense is because it's encouraged, and it's encouraged at some companies so that people will break out of their existing workflows to hopefully find useful new ways of working or building
There is more to it than this, but much of the cost structure around subscriptions etc is specifically designed to allow for that experimentation.
There are good cynical takes, here, too. At the current model costs I don't need to optimize my expenses, but that could change if it climbs eg above 30% of my salary^
Note: this is an easy thing to prove ROI on. If I'm writing 5-6x more code and reviewing commensurately more code, and those PRs are better-tested and get us to shipping quality features faster, this is easy to justify and we are not that price sensitive
My understanding is that distillation uses specific probing queries to extract model weights in some meaningful sense.
There's also the element of "how one uses model outputs can be in accordance with or against the TOS"
(This is true of any website! How I access and use the info on the website can be in accordance with or against the TOS)
Both of those things: doing something against TOS, using specific queries to illegitimately gain access to model weights - that's why calling it adversarial and stealing feel fair.
I didn't (and don't) understand your point, but in an earlier post you made some claim about it being Anthropic - you can't steal from Anthropic. The above is the only reason I can think one might believe that
> Theft presupposes a legitimate possessory claim by the victim. If A’s possession of X is itself wrongful because A stole X from B, then when C takes X from A, C has not violated A’s rightful ownership of X—because A has none.
I think the whole thing is a bit fraught. In the best case, all frontier model companies would have invested in a giant expansion of Wikipedia and thus distillation would be stealing because the base information is already public and available. Obviously that's not what happened.
However, at this point, I suspect the stolen books (and scraped websites) are largely a footnote of training. Something that was essential to create early models, but relatively minor given the work expended since to create new content and RL environments
This is the most credible-seeming claim about why a competent administration might suspend access (by any means necessary, but also by export controls) to models like Fable
However, I haven't seen any prominent articles proposing this theory, I haven't seen anyone in the administration gesturing towards this as the reason (but haven't been following too closely)
Do you have any sources?
(And in fact, it seemed like an obvious hypothesis that wasn't getting much air time in the first weekend, but again, I didn't see anybody really staking a claim to it except in a few comments or tweets like this one)
It seems like a reasonable person could say that a model being distilled "against the model provider's wishes" is in some sense a cyber attack that is stealing information (eg the lower order bits of the model weights)
I think this is mostly a confusing way to describe it, but I'm not really sure why you say it isn't an attack or adversarial. One side is doing something the other side doesn't want. Seems to be pretty clearly adversarial.
They are losing a ton of money on this and would be much much better off if people (Americans or not) had been building with Fable over the past few weeks rather than not
If Anthropic wanted to continue slow rolling their best model to a select few companies they could've kept expanding mythos access instead
Not to mention the fact that OpenAI has had time to prep 5.6 for release so that there will actually be an alternative around the same intelligence point as Fable. This is just all around terrible for Anthropic.
If you want a conspiracy theory, the one that says "this is corporate assassination" or "the US government does want to be in the business of picking winners" is much much much closer to the truth than yours.
Rather than buying either conspiracy theory, though, I expect Hanlon's razor describes the situation best (incompetence on one side, inability to coordinate/communicate effectively with the current administration on the other)
Whether or not you assume bad intent of the government here, the amount of money the export control on Fable has cost Anthropic has to be unbelievably high
Eg I expect I would have paid more than 2x per day what I spent the past few weeks, and if gpt 5.6 comes out and is competitive that's going to absolutely gain market share.
Right, but there exist problems that need to be routinely solved and can be solved on glm 5.2. is the model state of the art when it is published? No. But when it comes out you could optimize it and let your solver run forever for quite cheap, and that could be useful if the only problems you want it to solve (for cheap) are solvable by that model.
And the high water mark of what can be solved by open models will keep going up.
While I agree that anthropic has several communication and PR problems, it doesn't seem like Fable has been shown to offer any advantage here (for cyber offensive capabilities) over the previous state of the art.
I'm not saying all of Anthropic's statements are true, but mythos did seem to find many legitimate security exploits. You should be able to talk about a helpful-only model being released to limited partners while still releasing a very locked down model that doesn't advance the state of the art on these things, and that seems to be what they did.
How do you think founders should feel about advertising then? Eg Ben Thompson always argues that consumer products where you want to maximize audience (eg to distribute fixed costs, benefit from scale, power a multi-sided flywheel) should prefer advertising, but obviously this forces a specific trajectory to the business.
If you run into a product with this type of dynamic is your advice just to run? Or is there strategy innovation we can do here to monetize businesses shaped like this to get a flywheel without ads?
Suppose you own a thing. And it becomes extremely profitable to own. And so you're able to pay people extremely generously to help maintain the thing. So everybody you pay feels like they're getting a good deal by working for you.
And the thing you own becomes so valuable it's worth a billion dollars.
You are now a billionaire. But through your telling, you don't deserve it, and that might be right. We didn't discuss how this amazing thing came into existence, and if it just magicked into being then sure, you don't deserve it.
But suppose that without you, this thing never existed. How should credit be distributed?
Now there are lots of problems here and there are lots of ways to criticize startups and many of them are legitimate. Oftentimes companies do exploit their employees, or use exploitative contracts, or are exploiting some resource that we don't like them exploiting. But almost every conversation like this implies that this is the only way, that it's theoretically impossible for a company to do things legitimately. And honestly, that's often fair because corporations often become extremely extractive / exploitative (see the new book by Eric Ries if you need examples and counterexamples)
But I would like it if everybody could correctly realize: the problem isn't making a billion dollars. It's how we do it, and it's the incentives that we place on companies for continued growth. If you're a politician you should work to fix _that_, not the existence of billionaires.
> most people who’ve built something people love are not billionaires
I don't think his advice was simply "build something people love." In fact, he specifically spends a lot of time trying to make it extremely clear that a necessary ingredient is compounding of an extremely high growth rate.
So I'm not sure if your take is intentionally misleading or if perhaps we read different essays
I'm somewhat confused by this. Many of the early startups are literally just the founders, and if they find product market fit without bringing anyone on, there's nobody else to give credit to.
I've talked to a lot of seed stage startups this past year, several of them have achieved PMF and have several large customers. None of them have more than five people. If the companies didn't exist, nobody else was going to build the things (most of them) are building.
Where should you assign credit in this case?
Some of them largely eschew AI programming assistance as well.
Surely for these companies, if the founders get to several million or tens of million in revenue without hiring any more people, those people have successfully become millionaires and we can credit them as such, right? Or do you simply think this is impossible?
I expect ticketmaster has what should by any reasonable court be deemed illegal contracts that prevent venues from hosting non blessed artists and artists from frequenting non blessed venues, and this type of contract lets them maintain their monopoly
I think some of his advantage analyzing where tech can go is because he pushed the limits of it (eg working remotely early early).
He was disappointed in the Apple vision pro for just being an entertainment device (it seems like you two agree there?)
And then the interviews by media of tech should be viewed as an iterated game. He can ask interesting questions for an analyst, but he (and Nilay) do depend on access and that fundamentally constrains what types of questions they can ask if they want continued access
> Just not doing it for me. Think I'm gonna stop reading anything he says.
Unreal is like venture capital or a book advance (or the equivalent in music record deal)
Can you self publish? Sure, of course, have fun. But if you want the support and infrastructure of a company that understands the business of books, you take a deal and it is just like this: if a bunch of authors get book advances, that is generous to the ones who are unsuccessful, and they can only do that _because they capture the upside of those that are successful_.
Without that, you don't get advances for anyone.
So the point I'm making here: unreal provides variance reduction for all game publishers and yes that disproportionately benefits the ones who make under a million. But they're the ones who need the help!
And in exchange, if you're one of the lucky few, you pay a shockingly reasonable 5% in perpetuity.