margalabargala

Yes, that's accurate.

Are you trying to say "the Anthropic API surely cannot be highly marked up! Therefore subscriptions surely are selling at a loss!"? If not, I'm not sure how your comment is particularly relevant to the discussion?

Subscription plans are not massive losses. What are you on about?

They are simply less profitable than API pricing. They are on par with what sellers of open model tokens sell for.

It provides an outlet for people interested in finding opportunities to dunk on the US, allowing intelligent discussion to happen elsewhere.

> no doubt that Japanese lifestyle is much better than the American lifestyle

Smoking rates lend some doubt here.

It could encourage face to face socialization during those hours, which could improve mental health.

That's not necessarily the cause here.

While rates in the US may be rising, they are still below countries like Japan, Italy, France, and the UK.

Yes, doing releases this way means that it is confusing to compare model capabilities without going to enormous effort to run my own tests to determine something that should be knowable at a glance.

Waiting is the way to solve this. And despite the outcome being the same, making me feel like I'm being forced to choose to wait is a worse consumer experience.

If you judge everything by the worst thing it ever did, you will come to the conclusion that most things do not deserve to exist. Yourself included.

If you judge things based on whether they have had a net positive effect on the world, you'll find that both yourself and USAID are in the same category.

It's very different when written out, but when speaking aloud it blends and becomes difficult to distinguish.

I'm very much not from anywhere resembling the American south, and "you all" is very much a part of my lexicon.

Okay, now how many generations until you use Terra over Astra?

I think it is a problem, for OpenAI. If their answer to "which of your models are better" is to throw their hands in the air and say "fuck if we know, go figure it out yourself" then that's going to go poorly as soon as they try to actually sell to some of the market they're predicting they can sell to.

Whether they train the biggest model first or not is not relevant; what matters is what they release. Right now we've got 6.1 Sol and 6.0 Astra. Over at Anthropic they've got Opus 5.5 and Fable 5.1. The existence of a hypothetical Fable 5.5 that I'm not allowed to use doesn't help me.

You're misunderstanding the misunderstanding.

You're right! The relative sizing of models is dead simple. This doesn't help when there are improvements between release numbers, and models are released one at a time instead of as a family.

When Sol 6.4 comes out and the latest Astra is still 6.1, which do you use?

Yeah, their title is absolutely implying a GLM bug to anyone who doesn't read it super closely.

There's a huge amount of unnecessary extraneous details in the headline that distract from "this is a Cloudflare bug".

Yes, it's one strong stronger.