Apologies for the late reply Adam. I had more time to think about your article, and while I harshly called it science-fiction, perhaps it's just the percentages and dates that are fictional. And therefore the impact still a threat, but not as drastic.

If OpenAI's goal is to replace human coders, it must be based on eventually achievable metrics. Unless it was a publicity stunt/conversation starter, they will do whatever possible to bring it into fruition. Based on its current capabilities, they have a long way to go, but who knows the actual timeline.

The way I imagine it being technically qualified to take human jobs would not be based its ability to leverage fully public knowledge. E.g. how to mix proprietary information, business domain expertise, and software APIs to create something tangible. That would require each company to have their own privately hosted AI that they can teach their corporate IP to. I think there'd be so much red tape before that happens, both in general and per corporation. I think there'd be many intricacies to further that discussion, but I'll save my words and digress.

I believe that sometime in the future, some jobs will be lost. And perhaps later, more jobs. Though no matter how good AI gets, I hope there's still many jobs that will always be safe.

I think we also have several years to build up on skills to make us irreplaceable. Whether it's deep tech knowledge that AI cannot grasp, business domain, decision making based on wisdom and experience (not just logic), or just soft skills that have nothing to do with coding that make great programmers great.