Actually I am not sure I want to go into AI. It's fascinating to watch from the sidelines but it doesn't look like a lot of fun. In fact, programming as a career started becoming a lot less fun a decade or so ago — not nearly as fun as when I started in the 1990's.
Maybe I'm mistaken though, likely there are a lot of cool as-yet-undiscovered uses for AI that would be fun to explore. Maybe a few of those will come about and draw me in.
As a fellow oldie, in programming at least, what I’ve found helpful in keeping programming fun after 25 years is to focus on the outcome moreso than the programming itself.
Creating something cool is fun. The process is just work.
On one hand: this is a nascent field where there's lots of low hangnig fruit for picking, and enough uncertainty/hype to trick people into getting easy money.
On the other hand: AI startups don't make a lot of sense. The research and large data are the tricky parts. It can work; Google fumbled the ball and Midjourney as a startup managed to create a really good model. But then what? once the theory comes out, everyone else with a bunch of money swoops in and creates derivatives. It's great for consumers, but terrible for startups that are expected to shoot to the top and become monopolies like Facebook. It's also not a development problem where you can incrementally make improvements. An app or SaaS, you have features you implement them one by one to make it better. With AI, you could spend years researching RNNs for NLP only to have something like transformers come along and crush your dreams and make years of effort obsolete.
For me the keyword there is "pivot", not AI. My current expertise will continue to be needed, and having better understanding on how AIs work and can be used probably will be able to do my current work better.
Maybe even take a step forward and learn how to run AIs, performance, monitoring , scripting and other things that I do for other systems. Or some high level abstraction on how to feed them with specialized information.
But pivoting, as in leaving all behind and focus on just this, that doesn't seem a good choice at an age not so far from retirement.
At least, not at the current stage of development. If the whole landscape is totally different in a couple of years because we are reaching right now a technological singularity, the best option may be different.
What is everyone’s thoughts regarding switching in the first place? If you have switched, what lead you to that decision? I feel like software is a cool career in that we’ll never make “bad” money even if we more around. Even if we may never reach our full salary “potential”.
10 comments
[ 3.0 ms ] story [ 30.7 ms ] threadhttps://news.ycombinator.com/newsguidelines.html
"Okay, you are in fact too old."
Actually I am not sure I want to go into AI. It's fascinating to watch from the sidelines but it doesn't look like a lot of fun. In fact, programming as a career started becoming a lot less fun a decade or so ago — not nearly as fun as when I started in the 1990's.
Maybe I'm mistaken though, likely there are a lot of cool as-yet-undiscovered uses for AI that would be fun to explore. Maybe a few of those will come about and draw me in.
Creating something cool is fun. The process is just work.
What I hated about The Industry is they way so much process crept in.
On the other hand: AI startups don't make a lot of sense. The research and large data are the tricky parts. It can work; Google fumbled the ball and Midjourney as a startup managed to create a really good model. But then what? once the theory comes out, everyone else with a bunch of money swoops in and creates derivatives. It's great for consumers, but terrible for startups that are expected to shoot to the top and become monopolies like Facebook. It's also not a development problem where you can incrementally make improvements. An app or SaaS, you have features you implement them one by one to make it better. With AI, you could spend years researching RNNs for NLP only to have something like transformers come along and crush your dreams and make years of effort obsolete.
>Software 3.0
yeah
There's interesting ideas, and then there's revolving your entire product around a ChatCompletion API run by your biggest competitor...
Maybe even take a step forward and learn how to run AIs, performance, monitoring , scripting and other things that I do for other systems. Or some high level abstraction on how to feed them with specialized information.
But pivoting, as in leaving all behind and focus on just this, that doesn't seem a good choice at an age not so far from retirement.
At least, not at the current stage of development. If the whole landscape is totally different in a couple of years because we are reaching right now a technological singularity, the best option may be different.