

While true, the latest opus model has 1m token context. Which is a lot more than the previous 200k limit. Hard to fill that up with regular work, but easy if you try to oneshot a whole product.


While true, the latest opus model has 1m token context. Which is a lot more than the previous 200k limit. Hard to fill that up with regular work, but easy if you try to oneshot a whole product.


While I agree with this, and I’m not defending skirting regulations, before rideshare apps, taking taxis was an awful experience. At least half the time, if you try to pay with a credit card, the machine was “broken”, if you wanted to get a ride at a specific time you had to call ahead and hope that a taxi would show up.
Rideshsre apps forced regular taxis to up their game and provide better service, some did and now have their own apps.


Agreed. They have steadily added new features that don’t bloat the experience. It’s just a little too constrained for my tastes, but the rest of the Polish makes up for it.


Same here. It’s very smooth. I wish there were a few other customization points, I’d rather use my own weather app than the built in one, and I do, but then end up with different forecasts as they use different sources.
Minor complaints, but still a little friction that could be solved.


I mostly agree. It won’t replace people directly, but it might let companies do more with fewer developers.
Not to say they don’t need anyone, but maybe they get by with 75 instead of 100 developers? Hard to say where that falls because the outcomes if ai usage are so variable based on skill of operator and the target codebase.


Let’s be generous for a moment and assume good intent, how else would you describe the situation where the llm doesn’t consider a negative response to its actions due to its training and context being limited?
Sure it gives the llm a more human like persona, but so far I’ve yet to read a better way to describing its behaviour, it is designed to emulate human behavior so using human descriptors helps convey the intent.
Pushing SSL was probably the last big tech effort/push that actually benefited users. Sure it made self hosting a little harder, and probably consolidated some tracking behind bigger players, but overall end users did benefit.
Most of what I see now is purely for their benefit and users don’t benefit.