[awb]

It’s got to be one of the first areas to lose significant jobs.

Products are more about audience and experience rather than amazing technology.

If GPT4+ can spit out an API and an unstyled front end that a designer can beautify, that’s a massive productivity boost. It’s not there yet, but it’s improvements are amazing.

I find myself acting more as a software architect with GPT, using it to spit out functions.

But there’s a cultural shift that happens when your non-technical friend starts using technology. Like when non-tech folks started embracing the internet or using smart phones or creating social media accounts.

When everyone you know can use GPT to build software, that’s when you’ll be out of a job. Until then it’ll probably just become increasingly more competitive.

[oytis]

> Until then it’ll probably just become increasingly more competitive.

Will you? If your competition is becoming more competitive too, no one is becoming more competitive. Just the number of engineers required to cover the market needs shrinks significantly.

[awb]

That's what I mean: fewer jobs for the same number of engineers would mean greater competition for each job.

And given how much of an accelerant GPT is, I think learning how to prompt and validate responses will be increasingly important.

For example, if you compare an average CS grad with a 4y career to an average CS grad with a 6 year career, the more experienced developer might be 5-20% more efficient.

But with GPT, an engineer who writes amazing prompts versus might be 2-4x more efficient than an engineer that writes average prompts and has to spend time fixing code, debugging or re-prompting.