The models are noticeably different — for example, o1 and o3 have reasoning, and some users (eg. me) want to tell the model when to use reasoning, and when not.
As to why they don't automatically detect when reasoning could be appropriate and then switch to o3, I don't know, but I'd assume it's about cost (and for most users the output quality is negligible). 4o can do everything, it's just not great at "logic".
I worry that day jobs are useful far beyond money for side projects. You do what you don't want all day, but then you are able to feel happy enough to do great work when you go back home to do what you really want.
I actually enjoy my day job. It's just that it's a large part of my day and I feel that I my other interests (restoring old computers, exploring them and how people worked with them, mostly) end up being neglected.
I think it works as it should work. The problem here is the amount of content and how they're evaluating which video to recommend next. If these systems are showing things you'll likely spend more time with rather than useful things or things you'll enjoy the most, then I agree, it's got a fair share of anti-recommendation.
I suppose they do not have any metrics except what the user is actually seing. When I have understood this I have stopped watching some kind of videos I do not want to be spammed with in future. Also the following trick really helps to enhance the recommendation: watch these [1] videos in silent mode not for the sake of watching per se but for the sake of persuading the Algorythm that I am a science geek and not a tiktok teenager.
I built this thing a few months ago because I wanted to write ideas or whole stories down at times when I should be working instead. It's awkward to get caught writing things you don't want other people to see yet (even more at the office).
It's been useful at times, though I really don't use it that much. I'm just (secretly) sharing. Maybe it will come in handy for some of you.
The "Search notes on ChatGPT" button is just to make it look a little more like a "serious" service.
