I never had the impression that there were enough people for the amount of work anyways. I don’t see jobs go, but shift. Most developers will be fine, because of never ending work, AI is just a tool speeding things up. But not that much, as someone who is good with Google and git, is just a bit slower to find the same answers. And AI needs verification too, even if it links you directly to the issue at hand, via source url.
AI will create new issues. Some of the low level requirement jobs will go, like working in first level support, but only if you learn the AI yourself, else it’s too generic. We’re not there yet, where companies learn their own LLM yet. some outlier try.
We got to understand that there’s still a human layer and a lot of people might prefer calling a human, even if the result is worse, simply because we’re social beings. This can cost a lot of customers, if companies believe they can just shove an AI in front.
No one really knows how good AI will get. As the technology advances, we find more and more hard to solve issues, for instance that AI will make things up or gives wrong answers, despite knowing the real answer, if you pressure hard enough.
Also for security reasons you can’t add AI everywhere, unless you want to send all secrets directly to Microsoft, Google or Facebook.
I never had the impression that there were enough people for the amount of work anyways. I don’t see jobs go, but shift. Most developers will be fine, because of never ending work, AI is just a tool speeding things up. But not that much, as someone who is good with Google and git, is just a bit slower to find the same answers. And AI needs verification too, even if it links you directly to the issue at hand, via source url.
AI will create new issues. Some of the low level requirement jobs will go, like working in first level support, but only if you learn the AI yourself, else it’s too generic. We’re not there yet, where companies learn their own LLM yet. some outlier try.
We got to understand that there’s still a human layer and a lot of people might prefer calling a human, even if the result is worse, simply because we’re social beings. This can cost a lot of customers, if companies believe they can just shove an AI in front.
No one really knows how good AI will get. As the technology advances, we find more and more hard to solve issues, for instance that AI will make things up or gives wrong answers, despite knowing the real answer, if you pressure hard enough.
Also for security reasons you can’t add AI everywhere, unless you want to send all secrets directly to Microsoft, Google or Facebook.
My 5 cents.