foolswisdom

Seems to me that it matters because it has implications for how we can train smaller models?

For some types of applications, Jev is also useful as a stepping stone, you use it for the cases it's most confident so you already get gains from using it, and then collect all the usage to train your specialized model.

Brings https://web.archive.org/web/20260418203923/http://gamestudie... to mind.

Probably not, the network towers are the ones being used to triangulate I think.

Apple isn't the most dominant business on earth, so maybe they can't do anything in violation of antitrust?

Not how it works. You define a relevant market, and the market of iPhone users (which a stable market, there are barriers to just switching to android tomorrow) can be used for evaluation.

I was sure I had read of this year's ago on Wikipedia as the stable marriage problem (originally heard of it in a podcast), but a couple days ago I saw the wiki page said stable matching problem...

Maybe in "Iberia" instead of Spain?

> The State’s second argument concerns TikTok’s representations about whether its “Restricted Mode” feature effectively filtered out content that might be inappropriate for minors. TikTok claimed that Restricted Mode “limits the appearance of content that may not be appropriate for all audiences” and that users “shouldn’t see mature or complex themes,” such as “profanity,” “sexually suggestive content,” “realistic violence or threatening imagery,” “firearms or weapons in an environment that isn’t appropriate,” “illegal or controlled substances/drugs,” and “explicit references to mature or complex themes that may reflect personal experiences or real-world events that are intended for older audiences.” For years, the State argues, TikTok employees internally recognized that Restricted Mode was not functioning in the way it was marketed and exposed users to a high volume of content that should have been filtered out.

https://www.texasattorneygeneral.gov/sites/default/files/ima...

A 180 might be more useful :)

When it's happened to me, I shrugged and went back to the way I did things before. Then again, I'm not a vocal social media user by any means.

Seems this one might have been deprecated a while ago.

I'm not sure if I understand, you're saying that there is no such thing as epigenetic inheritance? In other words you disagree with the phenomenon discussed in this wiki page? https://en.wikipedia.org/wiki/Transgenerational_epigenetic_i...

As a thought experiment, there's a big difference between "3 companies share 70% of a market, while 30 companies sharing the remaining 30%", and "30 companies share the whole market, none being larger than 8%".

In the former example, there's far more pressure on the smaller companies, which is why in such a regime people expect more consistent quality from the big guys, and the smaller guys are often boutique in some way (=offering something unique in their service/product in order to be able to compete). Small guys have a hard time standing up to big guys.

On the other hand, in a regime where no one becomes that huge, like the latter example, there's still pressure, but the competition is fair, just like if all the companies were similarly big (except that you can't have that many big companies in a single industry, by proportion definitions).

Realizing this drove home why inconsistent antitrust enforcement appears to produce bad results. Because it leaves us in the state of the former example. It's also unfortunately the intermediate state of getting serious with antitrust enforcement.

You're talking about different measures / types of quality.

> Anyone who thinks AI is good with writing code that is hard to write for the operator, not due to lack of basic software engineering know how but complexity of the domain, either has access to models beyond what is available to the public or is completely lost.

They're arguing that most usage of resources would be by companies, who are presumably paying.

Agreed. The recent privacy kerfuffle with Grok CLI (plus the untrue "it's your fault" response, although they then changed the behavior) is a really bad sign for how your code and data will be the treated.

From the article

> It’s hard to collect enough data using invasive methods. Few people want a chip in their brain! But non-invasive methods are getting much better. The hardware is improving and getting cheaper, though I apologize for being vague about the particularities of our hardware.

In the original its meaning is being part of a greater unit.

Claude seems to do this all the time. That, the empty words and the increased self initiative of claude were all reasons I decided to just use codex (now merged into GPT).

So you're saying that it depends whether you're measuring "success at task X" (where in average everyone improves) vs comparative success (where people with knowledge can do far and away better at specific tasks).

Yeah, a common problem in cognitive bias research...

The renting out is a pretty recent development.

Related: https://news.ycombinator.com/item?id=48964185

If enough people/companies are on the bandwagon, you're afraid to push back and be the crazy one, potentially losing money. So you roll with it.

While the misuse of the term monopoly is annoying, that's quite the misleading comment. You're allowed to disagree with what policy is a good idea, but laws like the linked do exist, and they were seen as pro-competitive in their time.

Today you can consider that "just business", and therefore "part of competing", but the were laws with the intention of allowing/disallowing types of tactics. E.g., you can't compete by leveraging your volume (see attached link), so you have to focus on making your service good.

The assumption made by a these people is that if there are so majy companies getting that big, there must be general disrespect of these laws. And correlating with the undisputed fact that antitrust enforcement did change.

https://en.wikipedia.org/wiki/Robinson%E2%80%93Patman_Act?wp...

It sounds like you're focusing on the problems of running local models, or running models yourself, but I don't think that many people seriously expect near term improvement on that, it's definitely more just hopeful thinking there. That's not what I meant to address, and I also am in more of a "wait and see" mode.

But at this point we do expect that open weights _hosted_ options become feasible for the tasks they're using the frontier models for. And because of the lack of "legal monopoly" (intellectual property of whatever kind), they're way cheaper, not mention more flexible.

The launch of the tinker platform from Thinking Machines is an example of the "more flexibility" part that people want (and they chose to make their model open weights, maybe because this is the angle they want to push).

At this point I think it's realistic enough that the ball is in OpenAI / Anthropic's court to figure out how to respond to this threat to their business model.

That said, I think it's concerning that there are apparently only a couple of providers of hosted open weights inference, due to the complexities of doing so (per Dax from OpenCode's tweets).

While preserving problems is undoubtedly a natural incentive, I think Hanlon's razor applies here. Just today I was reading Competent Bureaucracy - Rebuilding State Capacity (<https://cdn.sanity.io/files/d8lrla4f/staging/cf7eedaf5d21d27...>) on the topic of agency structure promoting success (the author has done a nice amount of work in the past - e.g. https://www.statecapacitance.pub into this history of this topic).

> Take a task, any medium-sized task, decently scoped that you'd trust to give to Sonnet to finish without a hitch. Now give it to ANY open-source frontier model and watch them struggle and go in circles while failing tool calls and randomly assuming things.

Claude used to be much worse than it is now, just as bad the open weights models are. And the open weights were worse. The labs will also try to keep the lead, but at some point people start seeing real value from open models. Maybe you say they're not ready yet for medium tasks, but everyone sees the writing on the wall.

Yeah, that's what bothered me when I read it, too.