I really don't know what AI could do for troubleshooting. I think troubleshooting should be done by a live person because problems can be out of the box and not so systematic that goes exactly by the book. Some problems needs a person that is able to think out of the box to flow with the problem itself.. uh? Does that make sense? lol.. hehe
It is fully dependent on how the AI is trained. I feel there is a risk is with recursive AI training giving the appearance of depth and breadth, but it really isn't.
I'd love to use it but it requires connection to AWS or Azure hosted AI cluster... I'm on air gapped networks so it's not gonna worky. There are ways to have self hosted container that contains trained AI models in air gapped networks but we're not there... yet.
I have embraced my new BFF called AI! Troubleshooting with AI is amazing; I am training my own model right now to assist with my job duties, I am having the time of my life. She helps with documentation too... I thought I was old... not going to make it in IT as a senior, once I relinquished my identity to Google, the rest has been just plain fun!
Someone check for pods in zennifer's cellar.
But that is a big part of why I want AI. It can run through the data, linked some of it together and make the basic assumptions for me. Then I can decide if I need to look deeper and even where to start looking based on the evidence provided by the AI. The simple stuff it can handle, but the complex stuff - just doing the simple stuff first will save me time. Granted this is based on how the AI is trained/built, and what data (or how much/little) is provided to it. If I know what and how it will approach the situation, then I can take the next steps.
cool
