AI is one of those buzzwords, just like “cloud” in years past, where the word has almost lost all meaning and every new company seems to be “AI-driven.” But, it’s important that we don’t forget that there are some real and tangible uses of AI that can really help you make your day-to-day more efficient.
I’m hosting a livestream on the IT Leadership Lab next Tuesday and we’re going to be talking about AI. We’ll be talking to a couple of IT leaders on how they implement AI tools, how they avoid the “shiny new toy” syndrome that often comes with trending technology like AI, and specifically what tools they use that are actually helpful when it comes to running an IT organization.
Are there any tools that you use specifically when it comes to IT management that helps you run your organization better? Whether it’s people, technology, code, etc., are there any AI tools that you would recommend?
5 Spice ups
The only AI stuff I use at the moment is the simple auto-complete stuff in Visual Studio 2022 - mostly because it’s a natural extension of the non-AI counterpart that’s been there for ages (Intellisense), and it happens at such a small scale that it’s easily checked as I’m coding.
Kinda waiting for the dust to settle a bit more before I dig in to anything else, especially in the system/network admin side of my job, since I generally don’t trust AI (or at least the current incarnations of it I’ve seen).
5 Spice ups
molan
(molan)
3
Copilot and Claude.
I use both daily. Especially for PowerShell code generation, but also for other tasks like analyzing and summarizing log files
2 Spice ups
ajason
(aJason)
4
I just upgraded Jira Service Management to their premium plan, which included AI built it. I do like the ability to summarize comments on a ticket, which can help for tickets that have had a lot of work put into them. The general chat AI leaves a bit to be desired though - I asked it how many tickets certain people had submitted, and the first one’s answer was “sixty-four”. I then asked who had submitted the most tickets and the person it reported didn’t seem that active with the ticketing system, so I asked it to verify how many tickets he had submitted. “Eight” was the answer. So according to AI, 64<8. I’m hoping that improves with time…
3 Spice ups
I’ve found good use for Google’s NotebookLM product. It lets you upload large “sources” (PDFs, text files, single webpages) and then interrogate the LLM about what’s in the sources. I use it to upload large log files, my troubleshooting notes, etc and then I get much more accurate descriptions about what’s in the logs.
It can also generate a fake podcast between two AI hosts discussing whatever you want them to talk about in relation to the sources. “Hey boss, here is this 7 min podcast explaining the incident from last week…”. That part is kinda trippy.
4 Spice ups