Juniors
In reaction to the often-heard complaint that the SWE profession would suffer the absence of junior engineers due to AI + seniors doing all the work, I recently was thinking the following.
This is not going to happen, because there will always be parts of the job that seniors would like to offload to juniors. The difference is that the entire profession would require many fewer seniors AND juniors (though time will tell).
My other suggestion is that this is not going to be as lucrative as it was for juniors, attracting masses into the field just for fun and good money in perspective. That was the case since 2010s. Now, I think, this will clean up the field from occasional people. Only people truly interested in the craft would pursue the path.
I also think that learning for juniors would be compressed as well. The ultimate task of an SWE is to deliver a correct and error-free product. AI would do exactly that: reveal errors, technical and architectural debt much more quickly, allowing juniors to learn on their own mistakes. Exactly how it works now, just within a shorter time window.
Besides, in the organizations with good SWE culture, juniors would be saved from exploring SWE dead-end practices.
Competent Judgement Is Still the King
Finally, in general, the SWE profession - as it is defined by SWEBOK - has not changed for anybody.
AI is just another engineering tool with its own benefits and flaws. One of the thousands of tools already existing in SWE (yes, that order of magnitude - concepts, practices, approaches, packages, frameworks - the list is endless). It will be built into the profession naturally.
The ones who want the craft badly enough will stay and own the entire toolset.
And, as always, competent (!) engineering decisions and judgement will require knowing all those thousand tools. I expect the application of AI as the final "silver bullet", a replacement for all the SWE tools will fade away gradually, freeing up room for real SWE.
Would love to hear the opinions from your specific context.
