Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

What's the deal with all the random weekly quota resets for agents lately?

minimaxir.com

Post reply on HN
Print view View original

31–40 of 94 posts

Re: What's the deal with all the random weekly quota resets for agents lately?

#31

Post by somewhereoutth »

minimaxir wrote: post #12
dlenski wrote: post #6

This article reads like a description of gambling-addiction behavior; the author appears to be addicted to using LLMs. He's eagerly awaiting his next hit of dopamine from his favorite model. He's setting timers to be ready for when his next hit comes available. He's spinning up unnecessary queries just to start the timer ticking on new models. You could basically write the same article about some guy who sits in a ca…

Gambling addition implies dopamine hits from irregular and uncertain outcomes: as I note in the post, I don't get a dopamine hit from running agents. I know people say "LLMs are just gambling because they're next-token-maximizers that can't write real code" but with GPT 5.6 Sol (and a few new tricks I discovered) the outputs are much less irregular and uncertain. It's just typical engineering. "Not wanting to waste m…

> and a few new tricks I discovered

This right here. Any gambler would recognize that statement.

Re: What's the deal with all the random weekly quota resets for agents lately?

#32

Post by vitaflo »

serf wrote: post #10
dlenski wrote: post #6

This article reads like a description of gambling-addiction behavior; the author appears to be addicted to using LLMs. He's eagerly awaiting his next hit of dopamine from his favorite model. He's setting timers to be ready for when his next hit comes available. He's spinning up unnecessary queries just to start the timer ticking on new models. You could basically write the same article about some guy who sits in a ca…

> This article reads like a description of gambling-addiction behavior; the author appears to be addicted to using LLMs. you're projecting, it's just insanely irritating to work with a tool that 1) limits its' own use 2) with a random interval. keeping in mind that plenty of people are making money on token use.. this guy sets a timer to wake up for work; he appears to be addicted to work.

If you don’t want 1 or 2 then pay by the token. If that’s too expensive for you then now you know why 1 and 2 exist.

Re: What's the deal with all the random weekly quota resets for agents lately?

#33

Post by minimaxir »

somewhereoutth wrote: post #31
minimaxir wrote: post #12

Earlier quoted context omitted.

Gambling addition implies dopamine hits from irregular and uncertain outcomes: as I note in the post, I don't get a dopamine hit from running agents. I know people say "LLMs are just gambling because they're next-token-maximizers that can't write real code" but with GPT 5.6 Sol (and a few new tricks I discovered) the outputs are much less irregular and uncertain. It's just typical engineering. "Not wanting to waste m…

> and a few new tricks I discovered This right here. Any gambler would recognize that statement.

Said tricks improve the output in an objective measurable manner, not theoretical, vibes, or gambler's fallacy. (blog post forthcoming on that)

I've been researching LLM prompt optimization for longer than ChatGPT has existed; I was successfully optimizing the output of GPT-2 back in 2019.

Re: What's the deal with all the random weekly quota resets for agents lately?

#35

Post by trollbridge »

noworriesnate wrote: post #8
CamperBob2 wrote: post #7

Competition. If Anthropic had followed through on their plan to put Fable under API pricing, users would have jumped ship en masse to GPT 5.6 (or perhaps to K3 if it turns out to be good enough.)

Theoretically it'll still happen tomorrow. But we'll see--I think they've extended it twice?

Anthropic gave up. A $200 Max sub now gets you access to Fable 5.

Anthropic's game is over.

Re: What's the deal with all the random weekly quota resets for agents lately?

#36

Post by acedTrex »

Ive typed out several things, none of them really capture succinctly the level of disdain I have for people that this article is describing.

I cannot fathom the idea that one is warping their work on a given engineering problem around the availability of a magic next token predictor box.

Re: What's the deal with all the random weekly quota resets for agents lately?

#38

Post by perching_aix »

dlenski wrote: post #6

This article reads like a description of gambling-addiction behavior; the author appears to be addicted to using LLMs. He's eagerly awaiting his next hit of dopamine from his favorite model. He's setting timers to be ready for when his next hit comes available. He's spinning up unnecessary queries just to start the timer ticking on new models. You could basically write the same article about some guy who sits in a ca…

Oh this is a good idea actually, that reset checker site allows for setting up a scheduled agent job, so that GPT alerts me when there's been a reset :)

Re: What's the deal with all the random weekly quota resets for agents lately?

#39

Post by chaps »

minimaxir wrote: post #33
somewhereoutth wrote: post #31

Earlier quoted context omitted.

> and a few new tricks I discovered This right here. Any gambler would recognize that statement.

Said tricks improve the output in an objective measurable manner, not theoretical, vibes, or gambler's fallacy. (blog post forthcoming on that) I've been researching LLM prompt optimization for longer than ChatGPT has existed; I was successfully optimizing the output of GPT-2 back in 2019.

Respectfully, a lot of what you're saying in this thread sounds a lot like the lies that gamblers tell themselves. Saying this as someone with a strong tendency towards addictions.

Some of these things are only possible to really see in hindsight. Yes, you've been working on these things for a while, but these systems are notably different in their capacity and strings they pull on us.

Be well, please.

Re: What's the deal with all the random weekly quota resets for agents lately?

#40

Post by minimaxir »

dehrmann wrote: post #13
minimaxir wrote: post #12

Earlier quoted context omitted.

Gambling addition implies dopamine hits from irregular and uncertain outcomes: as I note in the post, I don't get a dopamine hit from running agents. I know people say "LLMs are just gambling because they're next-token-maximizers that can't write real code" but with GPT 5.6 Sol (and a few new tricks I discovered) the outputs are much less irregular and uncertain. It's just typical engineering. "Not wanting to waste m…

> dopamine hits from irregular and uncertain outcomes Like the LLM getting the solution right?

For the problems I work on with GPT 5.6 Sol and the checks and balances I have in place, I estimate:

- 80% of prompts get everything correct and are confirmed correct with manual validation

- 19% of prompts make a minor mistake based on an ambiguity of the original prompt (user error not LLM error), but then reliably fixed in a followup prompt

- 1% of prompts causes more problems than it solves and is more pragmatic to just revert

For 99% good output, there isn't much of a dopamine rush when there is good output. The dopamine rushes are for the From the other replies on this post, I suspect no one believes me, but I am offering these numbers in good faith.

Post reply on HN