1. The end of unlimited token subsidization and a new status quo of metered usage limits.
2. Adding to the above, self-initiated usage limit resets give you some control and again help lessen the sting of the end of subsidization.
Right now though, I think Anthropic and OpenAI are in open MAU war ahead of their respective IPOs and that might be a bigger factor.
So it's not that "limits went down"; it's that "the bonus offer ended as planned."
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 casino all day eagerly awaiting double-your-winnings bonuses or similar.
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.
I also have a Netflix subscription. I watch a couple of things on it and stop. I don't think to myself I need to maximize my subscription so let me watch movies all the time and wake up at 2 AM to make sure the next movie starts.
Do you understand how the psychological response to the "random" disappearance of an annoyance is pretty much exactly the same as the psychological response to the "random" appearance of a reward?
I put "random" in square quotes because neither are in fact totally random, but both are clearly quite carefully engineered to provoke the desired response.
> you're projecting
I am not projecting. My total lifetime gambling consists of maybe 10 or 15 cash poker games with high school friends, ten minutes at a casino in Montréal which I found a revolting experience, and receiving a few $1 scratch lottery tickets as party favors.
"Not wanting to waste money" is the polar opposite of gambling.
Like the LLM getting the solution right?
- 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 <1% odds.
From the other replies on this post, I suspect no one believes me, but I am offering these numbers in good faith.
I think many people who don't believe you just haven't built-up the kind of prompt history & MCP / CLI tooling etc that lets you get to the point where things work at that level of accuracy.
Hope it helps to know that at least some of us here understand and are seeing the same thing. And if it's anything like my experience with Fable, "always be more ambitious". The capabilities of the models are often limited only by what you're brave enough to ask for. I keep finding I'm not ambitious enough.
Your post literally describes your fascination with trying to figure out the pattern of a "random" reward that you get, and trying to maximize the value you get out of it.
I put "random" in scare quotes because I strongly believe that—just as slot machine payouts are carefully structured to keep you playing—these LLM resets are structured to keep heavy users like you coming back to max out their usage, and to progressively upgrade it.
Several other commenters have also stated this same suspicion about the pattern of resets you're describing.
> "Not wanting to waste money" is the polar opposite of gambling.
From everything I've read about gambling addiction, particularly Jay Caspian Kang, that seems wrong.
The desire to "not waste money" and "get back to even" seems like a huge part of what motivates gamblers to keep gambling.
As someone who had family members go through gambling addiction this is the primary mechanism behind it.
Addicts don't see it as "cool fun dopamine kicks" but instead find it the only way they can get back to normal/where they are supposed to be
Logically, gambling is like going to the movies. You expect to pay x currency for y value of entertainment. If y falls short of expectations you might feel like you wasted your money, but who becomes addicted to going to the movies to try to get even? There is probably someone who has, but I’ve never heard of it and it doesn’t seem to be common; not like gambling addictions. For all intents and purposes it doesn’t happen.
But gambling addictions do happen, fairly regularly. Perhaps it is loss aversion coupled with the aforementioned dopamine hit associated with gambling that makes it so prevalent?
People who are really into roulette get a buzz off seeing the wheel spin, hearing the ball bouncing etc.
Probably that comes after the initial addiction to the reward function but it then strongly reinforces it.
I don't think there's an equivalent for paying to watch a movie because the time to payoff (or not) is too long and the sensory experience is too inconsistent to elicit a conditioned place preference.
LLMs on the other hand... the time to payoff is shorter and the experience is consistent every time. It's just lacking the tactile/sensory elements
I don't think gambling is at all like paying a set price for a ticket and having a pretty good idea of how long the entertainment will last. If "gambling" means making a series of short-term bets for entertainment value, you don't have any clear idea how long you'll be entertained for or how much it will cost.
People will show up at a casino with let's say, $200 and a debit card, and expect that they'll be able to spend $50, be entertained for 2 hours, and then just leave… while secretly hoping that they'll actually leave with more than they came with.
Then they burn through their $50 in half an hour, and dip into their remaining stash in order to keep playing. OR they triple their money in half an hour and feel such a rush that they want to keep playing with more money. Then, repeat repeat repeat.
Generally you can gain a pretty good understanding of how long the entertainment will last. Maybe not down to the millisecond, but you know a spin on the one arm bandit won't take hours. It will give on the order of seconds. Your willingness to give up x is contingent on the perceived entertainment value of those seconds. If you choose to play again, that is an independent event — like deciding to watch a second movie while you are already at the theatre.
> while secretly hoping that they'll actually leave with more than they came with.
Yes, this may be the undefined variable. The dopamine hit of believing you can come home better than you started, with little tastes of the possibility, coupled with loss aversion when it isn't being realized. This is what earlier comments seem to be speaking about, so perhaps, despite your insistence, a consensus was already reached.
This right here. Any gambler would recognize that statement.
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.
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.
Every single prompt worked without issue, and it got most of the way on the first try with the initial prompt (+ a couple visibility bugs due to the agent not having Computer Vision to see said menu bar app) such as:
> Create a SwiftUI menu bar app named `swiftmote` using theto create the most user friendly app following Apple's HID guidelines for creating a remote that can operate a Apple TV on a local network. Instead of reimplementing the protocols needs to interface with an Apple TV, use the Python package `pyatv` and host it within the SwiftUI app as a sidecar along with a Python installation.
I have my own Apple TV I can manually verify that it worked as expected, which is notable because the agent can't test or lie about this pipeline because it does not have access to the Apple TV.
That is not hallucination or psychosis. If you want, I can release all the prompts I used. (EDIT: Sure, why not, here are the prompts. If I don't complain about something in a followup prompt, assume it worked correctly: https://gist.github.com/minimaxir/30fa820daa1392da13026ec6aa... )
Just -- do well for yourself. Deflection aint it.
That 'triggers a surge of dopamine and creates highly addictive habits' [thanks Gemini!]
LLM use for code generation does exactly that, sometimes it works amazingly, sometimes it fails inexplicably. Whether it is negative sum or not doesn't really matter. Indeed it may well prove to be negative sum, especially if we step back a bit and consider the business benefit of the code produced, not just lines of code or even features produced.
LLMs tend to be a bit random, but are still more consistent and predictable than slot-machines. Also, gambling tends to exert lower effort and higher dopamine hits than vibe-coding, making it way more addictive.
But LLMs are still addictive to some extent. Maybe its around the same level as other behavioral addictions like food, social media, or gaming addictions.
This one just makes sense though. If you have a body of work to do that you know will exceed your 5 hour limit, then sending a message 3 hours before you start so that the reset happens in the middle enables you to do a task in one sitting.
But setting an automatic prompt in the morning so limits are more sensible during the day and allow continuous work is just common sense.
Might be the case here, but just setting a timer and being generally hyped about something is not enough for that.
Very weird how convinced some here seem to be that addiction is involved while they apparently don't know anything about the diagnosis criteria.
Anthropic's game is over.
What I read on social media about people and these resets gives off literal worst kind of addiction vibes. I've literally seen people talking about "Oh I had an existential crisis without Fable/GPT-5.6"
These people legitimately need help, or alternatively a social life.
Maybe its different on my end because I just use a sub outside of work for fun stuff. At work its not my money so I don't really care. I go to work, maybe use these subs at home every once in a while for a fun personal project and if I hit the limits (I rarely even do) I play video games or hang out with my wife/family.
People are borderline tying their identities to these models it seems, and yet most people aren't even building anything interesting.
Yes, 100%. The author seems to be describing, quite lucidly, how he is getting sucked into a gambling-like addiction to these LLMs… although he seems unaware of the implications of that.
Most likely (hopefully) 99% of users aren't like that and they'll just log off if quotas get slashed
It's really hard to blame the druggies when there's multinational billion dollar companies manufacturing the drugs.
But alas, we're in the grift economy, so the marks are to blame for falling for the fraud.
But for standard productivity, I've never come remotely close to hitting a limit.
More like trying to prevent people from moving off of them.
The only people that these companies are really trying to limit are the tokenmaxxers on flat rate plans and enterprises on flat rate plans that are cheating and should be on the API plan.
For everybody else, it's way better to reset their quotas if you have the capacity. If you cut someone off, they're likely to either associate your brand with being unreliable or go try a different tool. Both outcomes are going to cost you money.
WHen it doesn't turn out that way, it opens up options.
At the same time if other model providers were anticipating Anthropic or someone else to have problems at launch, and were waiting in the wings with their own models to launch competitively, it sets off a capacity release competition.
One might lower prices, one might give it away, or just make it available, and then there are users on other platforms with limits, or free until a certain day, etc.
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.
Hurray market competition and what capitalism was meant to be. Go above in providing a service or the customer leaves.
My expectation is that this cash handout is going to stop soon, so take while the giving is good right?
I should note that for my workflow /loop is eating the credits, so it’s basically no skin off my back. I just queue up some more work and let it go.
It seems a huge "the sky is falling" to think all LLM use is bad for fear of some "addiction" to me. Even many skeptics (eg Hashimoto) have come around to the idea of using LLMs.
One thing is clear in my mind. VCs are burning cash, and so if you have an LLM flow you find useful, take some free cash. The sky is not falling in this respect imo.
Which is not a defense of AI, to be clear. AI may very well be a big societal problem, but in this context i don't see Hashimoto/etc becoming heroin addicts like ya'll are so concerned about. It's just a fancy autocomplete. The fear seems a bit over hyped.
edit: Ok, I have been reflecting for a bit. This article really irked me and I realize now it's because I see a lot of a former version of myself in it.
Anyhow, to the author, I implore you to take a step back and look at what your mind is doing. Humans love prediction, and we love optimization, I get it. But, you're effectively rewriting history, and holding yourself to an impossible standard of perfectly optimizing past decisions with future information. As a result, you're sucking the joy out of pursuit that you seem to love.
I know it's easier said than done to not have your mind work against you like this. But, I think you might benefit from taking a step back and asking what your actual goals are in the work you're doing, and what sufficient value to justify the $100/mo subscription looks like in terms of output or work. Playing the token consumption optimization game is a losing battle that has turned a free gift into a source of angst.
I think the main problem is the usage limits. You would never feel like you are saving money by watching movies on Netflix. If they didn't have limits, many might find themselves using agents less actually.
Depending on your work, it doesn't feel limitless - I've just upgraded to a $200 plan and I'm starting to see where I will even hit the edges of that, especially once resets and special offers start tapering off. But with a single AI question often costing $75 in API costs, I'm not at a point where I can just switch to API and not care about cost at all.
That means I need to carefully time the work to make sure I'm squeezing what I can into the 5 hour windows. Feels a bit like the days of mainframes and time slicing, when you had to book an allotted time for your workload to run.
https://www.cs.cornell.edu/wya/AcademicComputing/text/earlyt...
Why can't people see the alternative hypothesis, inference has huuuuuge margins?
1. They have left over capacity that will go unused otherwise. Most users have hit their limits before the end of the week. Might as well get some use out of the servers.
2. A backend change made migrating usage tracking hard, inaccurate, or impossible.
3. A usage tracking issue meant the usage wasn’t accurate anyway and they needed a reset to save face / avoid contract breach.
The addiction or marketing strategy seems too clever. Why explain something with cleverness that can be explained by laziness or stupidity.
we (industry) have reached the point where questions about productivity gain, use-cost analysis and similar get asked increasingly more often
by resetting tokens they are silently decreasing the cost and increasing the use (hit limit ==usefulness stops until end of period) they are effectively fudging any "simple" internal company studies in their favor leading to a potentially pretty bad surprise if this isn't noticed by the assessing personal
An alternative is that they bought to many GPU resources in relationship to actual demand and they still need to show that they are "highly used".
Either its is appreciated for anyone benefiting from it but fishy as which company which doesn't do something fishy gives out rabbats to already paying users without turning it into a advertisement/PR benefit???
Combined with the fact that China's also catching up with its open source AI (no i dont want to debate whether or not it actually is or if china is 'safe'; particularly not with anyone from the USA as i am canadian) and also everyone protesting against the giant data centres AND the UK giving up on digital IDs?
yeah. they're trying to get people hooked again.
Open-weight is free beer without the freedom. But LLMs are evil by construction and using them makes people stupider, so I suppose "free beer [moderately cursed]" is appropriate.