I mean, even then, they could increase the price per token, if they want to hand out fewer tokens for the price paid.
They could make this work like a prepaid SIM card, where you charge it with e.g. $10 and then you can use it until the $10 are used up.
Instead, they make it work like in-game currencies in scammy free-to-play games. Except that they didn’t choose a confusing conversion rate, for some reason…
Yeah, I imagine that they did try. But it’s not just the intentionally misleading announcement post, they also have 5(?) different subscription tiers, which get different changes from this. And one of the subscription tiers is actually called “Pro+”, so that does not mean “Pro and more expensive tiers” like I wondered. And they have this ridiculous intermediate currency to make things even more confusing.
Their offering itself is overly complex and confusing…
Man, they couldn’t have communicated this more confusingly, if they tried.
Which is a crime, by the way, when you sell it together with a product you hold a monopoly for: https://en.wikipedia.org/wiki/Tying_(commerce)
So, did they use AI tools to type “LGTM” 400 times or nah?
But yeah, I also find that frustrating. Management just looks at terrible metrics like PRs closed or lines of code produced.
It’s not even novel that you can produce terrible code very quickly. Decades ago, our industry learned that it isn’t worth it, because you suffer for it later. Now the game is altered slightly and management demands that we throw all these learnings out the window.
Well, base prices stay the same. They seem to just be billing more per usage on top of that…
If there’s no reason to hold the feature code back (i.e. its integration doesn’t break anything), then it’s much easier for development to ship the feature and disable it with a feature flag. Otherwise, you have two versions of the code, which means changes need to be integrated in both versions, which is largely just pointless busywork.
Yeah, the big thing is that management has no sense how little coding you actually do in a software engineering role. You spend so much more time understanding requirements, understanding how you can resolve roadblocks within your organization and understanding what the hell the code does that was previously written.
In particular, the last part is something that will most definitely take longer for vibecoded programs.
The code is often needlessly complex, because:
But you also just don’t have human beings that made all the detail decisions and can tell you why they’re important. In vibecoded code, all of these detail decisions are accidental and only ‘proven’ in so far as the given accidental state that the code is in, happens to not explode in reality. If you need to tweak anything about it, you’re completely blind as to what’s actually important and what’s just in there, because the AI figured, it’s the most likely thing to autocomplete there.