[kfarr]

I’m reminded of the difference between being efficient vs effective. So much of the example use cases I see people — including myself — using GPT for are unimportant short-term tasks that necessarily take away head space and time from long-term important tasks. Those long-term important tasks are the hard ones requiring existing application context where I experience LLMs struggling. If we’re not careful we’ll get DDoS’ed by the tasks that an LLM can complete at the expense of other tasks. Of course this may change in the future as things progress, but is my observation for now.

[fatfingerd]

I think I get you, though I've been thinking of it rather differently.

I feel like a lot of the evergreen hype in computing is framework, practices, etc, that try to break things down into a system where any junior could then just fill in each piece and of course this always collides with the larger context problems.

Once you get to a certain point with such a system, either you have been paying attention all along or you have no idea what you've made and how to deal with a real cross cutting problem and you get to the point where the systems promise is really irrelevant, you succeed based on actual expertise you supposedly weren't going to need.

With GPT-like AI around its current level, I feel like some of these systems for breaking down programming projects are going to face an actual test now that the junior engineers to do it are some GPU costs that could be run in parallel and won't have the usual heterogeneous resources problems of testing with a real project team.

I'm not really sure if any systems will survive (or something learned in the process will make a good one) but I feel like it would be a proof of a holy grail that is suppose quite important, and just the refutation of many systems is itself a major disruption to the field.

[DonsDiscountGas]

OTOH, chatGPT can make a good rubber duck if you want to talk through a problem. It's advice is often useless (which is fine) but occasionally it may say something useful.

[fallat]

I find ChatGPT is scary BAD for rubber ducking. You have to already have an idea of what's right and wrong to verify. It is insane how often ChatGPT is wrong when I ask it things. like 90% of the time it is wrong. It's probably because everything I ask is way too specific and because it's just pattern matching on roids and not reasoning, it's impossible for it.

[kfarr]

Agreed you have to be very careful. The worst case I find very often is hallucination of a library for JS that either doesn’t exist or methods that are completely fictional. My initial response of wow that’s a perfect solution turns quickly into wow what a waste of my time.