Comment by al_borland
4 days ago
Our last CIO bought into all the AI hype early on and pushed it hard, before we were even allowed to use it. It was an odd time. We’d go to a town hall and get told that AI is going to change everything, and then get an email telling us we weren’t allowed to use it.
Now that we can use Copilot, we have a new CIO and I don’t hear about it so much. There is still some AI hype, but it’s more about how it’s being used in our products, rather than how to use it internally to do the work.
Apparently sometime in the next year we’re getting a new version of Jira with some AI that can do user stories on its own, but I don’t see that changing much of anything.
The bottleneck has rarely been the actual writing of code, it’s been people making decisions and general bureaucracy. AI isn’t solving that. Copilot has also not impressed anyone on my team. As far as the code we work on, it’s pretty bad. There are a few niche things it helps with, mostly writing queries to pull values out of complex json. That saves a little time, but hardly 30-50%. More like 1-2%.
Management stopped giving us new people, while pressuring us to do more, for many years now. This was a trend long before AI and I haven’t noticed any major change. I’d say it’s been this way for over 10 years now, ever since they had the realization that tasks could be automated.
For most senior software engineer roles at companies I've seen, the act of actually thinking about, designing, and writing code is probably 25% or less of your job. There's so much other non-development stuff that goes on. Advocating for features, writing boilerplate docs, seeking stamps of approval for this and that, reviewing other people's work, navigating internal process (your release is gated by Deployment System X which is waiting for Privacy and Legal sign-off), interviewing, writing down what work you've done for your next annual review with your manager, and of course meetings, meetings, meetings. Even if AI does speed up coding by 30-50%, that's maybe 10-15% of actual savings.
same experiences. lots of talk but most AI was locked down -- for good reasons, probably -- and we have explicitly allowed platforms.
in practice copilot is only useful for advanced search and it is really only useful for surface answers. Was a time saver for researching common tech I'm not strong in, like if certain Cisco platforms support X, or if there are simple ways to do Y in MongoDB, etc.
for example we had it search for 9-10 CVEs to see if any were exploited in the wild. Copilot got most of them wrong. ChatGPT got most of them right... except for one or two. But now I'm not sure I can trust anything and I'm checking them all from scratch anyway.
Both were confident in ways my security intern was not, and my intern was able to tell me "I have no idea about [CVE]".
Some people really bought into the meme where AI was going to replace everyone and only those aggressively pimping it would be able to keep their jobs.