The difference is, Devops isn't a bubble that everyone is waiting for to pop. I've been in that field for over ten years now, and properly implemented it is a net gain for everyone who does it. The reason companies are falling over themselves trying to hire 'Devops' is because they still haven't properly cottoned on to the concept but are afraid of falling behind. And yes, I can absolutely attest to the fact that Devops is a tough market to hire in at the moment, that there are a lot of places who don't have the first clue about what Devops really is, and - similarly to Agile - think they can add some buzzwords to their toolchain and call Bob their uncle. And there are a lot of candidates who somehow acquired a Devopsy title in all that chaos, but all their CVs have are tech buzzwords, and when you interview them they're clueless. That doesn't change the fact that Devops is a solid concept with high benefits for those who understand it.

AI, and more specifically GenAI and LLMs - is more like crypto, in the sense that people are trying to get rich from it without having the first clue what it is. It's this shiny new thing that everyone is rushing to get on board with, but I have yet to see someone propose a use case that actually makes sense, couldn't be implemented better without AI, and is a net gain for those using it. Right now it's all this nebulous bullshit, everyone just slaps their own coat of paint onto ChatGPT and calls it a day. Useful AI-adjacent concepts like Big Data and Machine Learning have been around for much longer than the tooling underpinning the current hype, and already have a lot of very valid use cases.

By the way, I work with a bunch of high aptitude Devops engineers and none of them are thinking about adding AI to our pipelines, not even to pad their CV.