ceenoteFeb 27
Thinking a lot of new technologies are stupid and unnecessary. Of course, at least in big tech, lot of them objectively are.
Show threadtuckermFeb 28

Yeah, this is me too. Although, to be fair to my younger self, back then a lot of new technologies actually were notable improvements over the previous tech, and older people were missing out by not trying them. I’m talking about going from cassette tapes to CDs, things like that.

Nowadays the new thing really is just worse than the old thing. E.g. going from a desktop environment to “the metaverse.” Those of us who didn’t embrace the metaverse were not just sticking to our old cassette tapes; the metaverse really was stupid as hell.

Show threadTjaFeb 28

The meta verse has 500 users, total, worldwide. Out of 8 billion people. It’s not exactly a popular new tech.

AI on the other hand, fits the description much better. It is very useful, and people who reject it out of spite are missing out. Sure, it’s being shoved down our throats in literally every product, whether it makes sense or not, I get the sentiment. But rejecting it completely makes you miss it’s useful applications.

Show threadnomy
What’s your use case thst you find it useful?
Feb 28 at 12:42pmWeb
Show threadTjaFeb 28
Boilerplate code, translation, text grammar correction, taking and summarizing meeting notes, debugging of network issues, interview role play…
Show threadnomyFeb 28

I have some experience but my attempts at using it for network debugging have been less than iimpressive. It’s able to give a great history/summary of the issue but when it comes to generating troubleshooting steps or an actual resolution it just spits out a one-size fits all generic answer.

Our attempts at using it for summarizing meetings haven’t been particularly great either but that may just be very boring, repetitive meetings lol.

Show threadTjaMar 1

Gemini in particular helped me greatly debugging a Thread (IPv6) issue with RA and SLAAC, giving exact commands and diagnostic requests.

For meetings work really well, I can concentrate on the content and not on taking notes. Of course if there is something really critical I will note it down on a paper. Then I need to spend just 5 mins reviewing the AI notes to make sure nothing is hallucinated, generating a summary and sending it over. What used to take me maybe 30 minutes now takes 5. With 4 to 6 meetings a day it add up fast.

It also helps my writing in foreign languages sound more natural, suggesting changes and expressions that I know exist but they don’t come naturally for me.

Anything “supervised”, to prevent hallucinations from doing too much damage. I save 2h per working day, easy.