> At some point the subscription model is going to become unsustainable for the frontier companies to continue (we just saw that happen with GitHub Copilot), and they will move everyone to a pay-per-token model.
From what I understand, Enterprise (above 150 seats, I think?) already has to pay per-token pricing.
Subscriptions are the premium "free tier" marketing of the AI world, so that employees can collectively request their large enterprise to subscribe to Claude, Codex, or Cursor, and presumably be billed at per-token prices then.
I'm surprised to no longer see Opus 4.6 on Cursorbench. I think there is a subset of Claude fans that are still adamant that Opus 4.6 is the best version.
> Then Sonnet/Haiku are just attempts to quantise/distil down to an acceptable performance/cost ratio. The cynic in me says we probably won't see any more of those until post-IPO, keep people addicted to the most costly models to pump a quarter or two of revenue figures, unless a competitor starts seriously undercutting them on price/performance. Hence the recent requests to slow down model training worldwide with their competitors.
I'm guessing there'll be a Sonnet/Haiku 5 release just around IPO, to keep the news cycle going, and so that user numbers will get a boost.
> And, I even use `claude -p` pretty regularly for scripted stuff (automated security vulnerability searches), which I thought was now counted at regular API rates, but that doesn't seem to ever run out either.
(Couldn't find a direct Claude link that wasn't a Twitter tweet (what is with AI platforms making announcements on that fElon-owned platform?!), so the Zed one seemed best.)
Ah, that explains it. I guess I'll see it show up as separate usage pretty soon. Or, I'll find an alternative for that role (though Opus is clearly better than DeepSeek or MiMo or any other cheaper large model I've tried).
I think Wikipedia's still considered unreliable, but the question that should be asked is whether the author even read the source in "the number in brackets" to ensure that it's even backed properly.
Just like how people should use AI for research, I guess.
Imagine that this website has a million visitors but just 100 rabid fans of one position. Imagine you read a comment. The website UX does not allow you to differentiate whether this is a person who is obsessive about one position or not. It doesn’t tell you whether you’re going against the consensus or not. So a small group of 100 users could create a bubble of visibility of a certain position. They could ensure you’re always voted down when you express a position.
You would never know.
The voting ring mechanic is hard to block but the comment mechanic is easy. Block a few hundred users and suddenly this site starts having much higher SNR.
This is just human behaviour though. We're wired for "a lot of people who do X, also do Y". "this person does X, therefore they must do Y". Obviously, not all brown things are cows, but that's how it be, it's got nothing to do with ai.
Right, but is there a difference between searching, say "acetaminophen and ibuprofen combined in emergency department settings" on google/ddg and asking an AI to give me primary sources for the same - if i am going to use the primary source anyhow? I just mention "i used AI to find this" because usually there's no good way to do a google search, or there wasn't the last time i tried.
For example, is glyphosate the active ingredient in roundup? there are studies that suggest not. I can't remember the university, i can remember the rough decade (2010s). all i know for sure is that someone showed that glyphosate isn't the active ingredient, really.
Deepseek can't find it. ddg doesn't come up with it immediately. I might try "deep think" mode on some other AI later, or use an older LLM model i have locally to see. I have the pdf, i just didn't rename it to be searchable! doggone it.
LLM assisted search is now one of the best ways to look into dense and obscure topics though, particularly given as google search quality has degraded. All it needs is for you to read the sources.
Source hallucination has also come down tremendously.
From what I understand, Enterprise (above 150 seats, I think?) already has to pay per-token pricing.
Subscriptions are the premium "free tier" marketing of the AI world, so that employees can collectively request their large enterprise to subscribe to Claude, Codex, or Cursor, and presumably be billed at per-token prices then.
reply