There's so many things that make me question this one:<p>1. It assumes company structure will stay the same. It won't. If 2 engineers and 1/2 a QA can do what a full 10-12 person team does... I can't see a manager leading a 3 person team.<p>2. If a structure like what is proposed takes hold, I'd be scared things would get too fragile. When you only have one expert on something, bus numbers become very real.<p>---<p>So, what do I see:<p>1. More cross project work. Because AI can help you understand new code bases faster, working on another project should become more common. Decreasing business risk.<p>2. The IP issues around AI must get cleared up. Especially GitHub Co-Pilot.<p>3. Code reviews for trivial stuff will be much more automated. That said, the human element is needed. Thus why I think teams will take on more projects. Specialization will decrease.<p>4. The business skill needed to pilot the systems at these levels will INCREASE.<p>5. AI will have an impact. But it will be different than we expect today. I don't think I can predict what AI will become. But I know it will change. To compare it to a human, I'm not sure how old it is... but it doesn't even feel like it is a teenager yet. It might still be in grade school to be honest.
Those blunt predictions stated with ignorant certainty and without any admission of doubt.. even already stating things like all development is frontend/backend/devops.. arghh.. cannot read anymore. Sorry (: (though must be great if the world is that simple)
It has been discussed widely on this site in great depth, but they've been predicting the end of "ops" for as long as I've been alive.<p>I'd pose a question to this author:<p>If not ops, who do you think will maintain/debug the AI infrastructure? Validate what it spits out?