I'm left wondering what AI everyone are using. I can prompt copilot and it gives me exactly what I need. Sure, if I barf out a lazy, half-baked prompt it yields a waste of time.
My problem is running into it's limitations, mostly around resources. I have tried giving it larger tasks and it takes bloody forever.
"Given this unstructured data, create CSV output for all platforms, with each line containing the manual, and model, ignoring the text in parenthesis."
Works great except for God-awful performance and stopping half way through. I had to break out each section and paste it into the prompt and let it work on small pieces. We need to get to the next level with this, especially for paying customers.
More concerning is that I see a clear pattern in smaller companies of hiring seniors and turning them loose with AI assistants instead of hiring junior devs. The prospect is attractive to nearly every stakeholder and the propensity to put off hiring "until next quarter" in light of this is a constant siren song. There is a lot of gravity pulling in this direction with the short-term thinking and distractions that are thoroughly soaked into the business world these days. Supposedly, one third of Gen Z (20-25 yrs old) are sitting at home, up from 22% in 1990.
I'm one of those seniors happily putting off hiring, but I find the situation and it's wider impact on the future very unnerving.
Well, having AI transform some data into a certain CSV format is orders of magnitude simpler and more straightforward of a programming task than what I try to use it for.
A lot of the discrepancy between people's experiences is simply due to the fact there's there's a massive range of programming complexity/difficulty that people can be trying to apply AI to. If your programming is mostly lower complexity stuff, non-critical stuff, or simply defined stuff, it obviously works better.
I try to use AI when I get stuck on a hard problem/algorithm, hoping that it can provide an answer/solution to unblock me. But when I'm stuck the problem I'm facing is so complicated that there's no chance at all that AI is actually going to be able to help me with it. I see absolutely no point in using AI when I already know how to solve a problem, I just solve it. I only turn to it when I need help, and it can never help me.
My problem is running into it's limitations, mostly around resources. I have tried giving it larger tasks and it takes bloody forever.
"Given this unstructured data, create CSV output for all platforms, with each line containing the manual, and model, ignoring the text in parenthesis."
Works great except for God-awful performance and stopping half way through. I had to break out each section and paste it into the prompt and let it work on small pieces. We need to get to the next level with this, especially for paying customers.
More concerning is that I see a clear pattern in smaller companies of hiring seniors and turning them loose with AI assistants instead of hiring junior devs. The prospect is attractive to nearly every stakeholder and the propensity to put off hiring "until next quarter" in light of this is a constant siren song. There is a lot of gravity pulling in this direction with the short-term thinking and distractions that are thoroughly soaked into the business world these days. Supposedly, one third of Gen Z (20-25 yrs old) are sitting at home, up from 22% in 1990.
I'm one of those seniors happily putting off hiring, but I find the situation and it's wider impact on the future very unnerving.