There has been considerable fear that developers will get usurped by GPTs. Thankfully, I became useless as a programmer in 1998 and never looked back.
Statistically speaking, my time as a "coder" was exceptionally brief. I got interested in the web in late 1993 and started hand-coding web pages in HTML a few months later. In 1996, I learned SQL over a long weekend in order to take a two-day class on Illustra's Web Datablade tool. In the process, I built one of the web's first content management systems.
The closest I ever really came to programming was Perl, which I also learned in 1996 from O'Reilly's Learning Perl (aka "the Llama book"), and then I promptly forgot it after I finished whatever project I was working on. I learned it again in 1997 and, yes, forgot it a second time, because, well, Perl. And it's not even strictly programming. It's more scripting than anything, or so I was told back in the day.
By 1998, the writing was on the wall. IGN had a dedicated database developer in John Windberg, who could normalize circles around me, and two front-end developers, Ilen Zazueta-Hall and Genevieve Buckley, who didn't appear nearly as confused by Javascript and CSS as I was. By the time IGN spun out of Imagine Media, I was no longer a programmer. I was a "product" guy.
Losing the ability to directly code wasn't the end of my software career; it was only the beginning. My job became envisioning the product, communicating the specifics to developers, and getting the hell out of their way. That worked pretty well for MyBlogLog and Gnip.
These days, I'm not even a product guy. I'm a "business" guy. I've coached some clients for *years* without looking at their product, because most startups fail from a lack of market understanding, not crappy UI. And, even still, people occasionally refer to me as a tech guy and it makes me laugh. Not even close, yo.
I bring all this up to say that smart (or in my case, passionate) people always find their way. ChatGPT might take the grunt programming work, but there's still the tricky bits to solve. And, someday soon, it will probably do the tricky bits too, but someone still needs to tell it what needs to be built in the first place. And, who knows, maybe we'll eventually just ask AI for software that does a thing, but, still, someone will have to know why the software is needed in the first place.
What programming looks like is evolving. What creation looks like is remarkably consistent. Sometimes we get help from tools. Sometimes we get help from other people. No matter what, we're problem solving and rarely doing it alone.
I'm excited as hell about what AI will be able to code next. I hope that all my programming friends and acquaintances are too, because they will always be needed.
#ai #coding #startups #evolution