Some of the technical knowledge I've picked up over the years feels pretty arcane now. With AI making answers so easy to generate, how do you keep the underlying troubleshooting skills sharp and make sure the answers are actually right?
Anyone else in IT feel like a fossil with all the AI tools?
Answers
Use AI to generate possibilities, not to decide what happened. Take each suggestion and ask what evidence would support it, what would rule it out, and how you can check without making the problem worse.
For example, a successful ping tells you something about ICMP connectivity at that moment. It doesn't prove the application is healthy. An unanswered ping doesn't necessarily prove the host is down either. Check the relevant service, port, logs, and network path rather than stopping at the first result.
Reply to this answer
Curiosity matters more than whether someone learned on old systems or new ones. The useful habit is asking why something works, then being willing to dig until the explanation fits the evidence. You can lose that habit at any age.
Reply to this answer
Being the person everyone calls when things break can feel good, but it's also a trap. If nobody else understands the systems, you're stuck being the only one who can fix them.
Walk someone through your reasoning during an incident or a follow-up. Write down the odd dependencies and the checks that actually helped. That gives the team something more useful than a collection of commands to paste.
6t~circuitbadger_429:
And it means the experienced person can take time off without the whole team waiting for them to return.
Reply to this answer
Old knowledge isn't automatically useful, and new tools aren't automatically bad. Some obscure details still matter because the systems are still around. Other stuff is just history. Keep learning the current tools too, including where their answers tend to need checking.
Reply to this answer
Add your answer
Answers are reviewed before publication.
8a!fiddlecrank_591:
That's the distinction I worry about. Getting an answer is quick, but knowing what the answer actually proves takes more work.