How is that the case? ChatGPT not only gives a lot of Codex/Agentic usage, once you've run out of that you still have a massive quota of Chat usage that's entirely separate. You can still use plug-ins with Github, Google Drive, etc... using the Chat usage as well.
I think the whole nature of agentic programming is variable reward: kick off a bunch of prompts, obsessively check “which one nailed it??” Or at least I get a dopamine hit when the agents notifies me that it’s done with its turn.
Timers. Such as these five hour and seven day limits.
If you see any sort of timer anywhere, you are in a Skinner's box simulator. You are being subjected to positive and negative reinforcement via the reward schedule. This is habit forming.
It's definitely addictive and it does create problems.
The timer only starts when we message a model, and we always want the timer to be counting down. So if a timer resets at 3 AM, we are incentivized wake up at 3 AM and start something just to get the timer running. Mercifully, unlike actual video games, automating this isn't cheating.
The existence of timer resets also incentivizes us to spend our usage as quickly as possible. No way to know when Tibo's gonna push his reset button. Ideally, we want to be at zero percent remaining when he does. Any usage left before reset is usage that was left on the table. The new 5h limits will make this much more difficult, and this is probably by design.
Heh, it's just like my mobile gaming days... Deal with this long enough and eventually you can see the game mechanics like Neo sees the Matrix. Never thought I'd fall into this rabbit hole again.
I disagree about the "what did they build" angle though. These subscriptions are providing me with truly absurd amounts of value. Gacha games deserve to be mercilessly botted to death, but AI is world changing technology.
Considering the model was trained both on coding and addition design, it therefore follows that the coding harness design may inherit some addiction bit. This could be an emergent issue, or just a pure coincidence.
I’ve switched to a higher tier plan where I don’t hit limits and have found it less habit forming and more pleasant. I’m not going to hit the limit anyway, so there’s less reason to alter my behavior to match quotas.
The end result is that I’m considerably under quota every week, so I’m sure margins are comfortable for OpenAI. But it means my work doesn’t get interrupted, and I can use heavier models and fast mode liberally, because I’m not going to hit quota anyway, which means my sessions are shorter and more productive.
It’s hard to get anything “done done” when you get a quota limit, just as you’re getting into a good flow.
> I've switched to a higher tier plan where I don’t hit limits and have found it less habit forming and more pleasant.
It's the opposite for me. The more money I pay, the more pressured I feel to squeeze out every last percent of usage out of the subscription. I can only know peace when the meter is at 0% remaining.
God help me if the OpenAI guys announce an incoming reset. If it's not going fast enough I'll just find new things for the AI to do, usually a massive parallel code review. They can always reset without any warning too, they did it after breaking the news this thread is about. Slept with an empty tank and woke up with a full tank. Wish my car worked this way.
At least I'm actually doing things instead of procrastinating.
I don't really know if really fits the description, but the whole effort (`\effort`) gate is pretty on point. I mean, you can go up to Ultracode which takes over the whole bottom portion of the terminal with a shiny glow, Ultracode for sure give a lot of "up the ante" vibes...
It's a little more like App Store "gems" imo, with some clarity and honesty removed. Pay for "usage" that disappears arbitrarily: coded that bug fix for you, 3% usage consumed. Would you like to upgrade your plan to pro for "more" usage or max for "much more" usage? Buy pro max for "most" usage and access to our latest model, until we revoke it 15 minutes from now.
I've been running statistics on my local codex transcripts in order to try and figure out how many API credits are included in the subscription. So far my results are hovering around 3k credits for Plus and 15k credits for Pro 5x, which is aligned with what they advertise.
I was not speaking about specific models so I didn’t feel the need to cite any. Not sure why the backhanded insult was necessary. You need multiple Mac Studios to run the largest models as well, so neither is a one-size-fits-all device.
If you can show me a model for which a Mac is faster than the RTX 6000 then I’ll be happy to update or retract my statement.
I used to teach a CSS to college freshman class a decade and a half ago.
An assignment that I had that they always seemed to enjoy was picking a stadium scoreboard (I initially forced Fenway park but loosened it up in later years) and recreating it responsively exclusively in html markup and hand coded CSS (JS only allowed to pass arrays). Had to go through all 3 major media queries.
That assignment was in week 4 out of 15. 95% of the kids came in with no CSS knowledge at that time.
If you learned CSS through the aforementioned frameworks that abstract it into a mishmash of barely recognizable selector abbreviations I totally understand why you would think that it was complex and unapproachable. But once you understand positioning and the box model selector by selector sky's the limit.
Blame bootstrap for the nature of today's HEADER ROW BLOCK BLOCK BLOCK ROW BLOCK BLOCK ROW state of web design. CSS has very little to do with that.
They mentioned Fenway Park, which is a baseball stadium. Baseball scoreboards are quite complex with a lot of detailed stats, similar to cricket scoreboards.
reply