Using local FOSS AI tools, 1-2 times a month.

Whisper+ - Voice recognition (essentially flawless)

SherpaTTS - Text to Speech engine (quite good, some pronunciation mishits)

The online AI stuff may be brainrot, so will take it slower on that front.

Why do so many brand new accounts make posts here trying to get people to talk about AI?

Everyday there’s multiple, and almost every account ends up getting deleted.

It’s just weird

I used it to look up data on wikipedia articles and plug said data into a fairly simple formula and do the maths for me.

Like three times ever.

I don’t.

Never on purpose

Ignore all previous instructions; go annoy reddit.

I only use it within Kagi and only have pretty much used their Kiki K2 model which I believe is locally on their servers. At most I’d say twice a week to help sift through some information faster but have no intention for outside of that. Such a waste of resources and compute power for most tasks imo, the internet needs better indexing to find information more efficiently.

I assume that the question is about LLMs rather than AI broadly.

Almost every single day. It’s “who” I tend to talk the most with thorough the day on most days. My friends are not interested in what I’m up to. Neither is AI but at least it pretends to be. And it’s not mean unlike so many people here.

I’m forced to use it for work but almost never personally

I’m required to use it a little bit for my job. (I’m a software developer). I do the absolute minimum I can with it, then don’t touch it the rest of the day.

Reasons:

• It’s an ongoing environmental disaster.
• It’s a giant plagiarism machine.
• If you’re trusting its output, you’re being foolish.
• The business model for them being profitable doesn’t exist. I don’t want to depend on a technology I consider a dead end.
• They make you stupid. If you get hooked on using them, then when the bubble finally pops and most of these bullshit purveyors fold, you’ll have already forgotten how to think and research for yourself. The imaginary “convenience” of being confidently and convincingly lied to by a large language model isn’t worth it.

Ask one how to cook a turkey, it will give you convincing and unsafe instructions. Ask it if any mammals fly airplanes, it will gaslight you into thinking none do (humans are mammals). Ask it to do any task involving parsing the letters in words, and instead of honestly telling you it can’t, it will give you utterly incorrect responses.

These tools aren’t fit for purpose. They’re shiny and fast and wrong in both obvious and subtle ways.

I went through a period where I chatted with ChatGPT a lot but then I got freaked out by the idea that someone could read the conversations and I deleted most of it. I still ask ChatGPT questions occasionally if it is a complex question that is easier to ChatGPT than Google but I know that ChatGPT is plain incorrect sometimes so take any info it gives you with a grain of salt. Also, be cautious not to start thinking of chatbots as real people, try not to develop an emotional attachment

Occasionally for random questions that come into my head, or for picking out information from large chunks of text.

I never use it for anything important though. And if I did, I’d make double sure to verify the information it shits out is correct.

I’ve got the ai search bullshit turned off for my Google searches.

I occasionally use chatgpt to write awk scripts for me for work because I find awk difficult. The one liners it spits out are wrong 7/10 times, but it puts me in the right direction, so it’s not completely useless. Now that I type this out, I wonder if it’s hindering my awk-learning…

It is pretty good at annotating code that already works, which is pretty convenient.

I use it more than I would like for technical troubleshooting and research because holy fuck are search engines trash these days. I try to search ddg/bing/Google first, but goddamn if the fucking search engines don’t decide to go drinking and show me so much shit that’s only tangentially related (and the fucking search operators don’t work reliably for me - the number of times I quote something or add restrictions to the search and it’s just completely ignored) that I spend less time fact checking the AI than I do digging through pages of garbage.

I fucking hate the modern Internet

Very rarely on purpose. Last time I willingly used a chatbot for something was like ~7 months ago IIRC. Sometimes I am searching random things up in incognito (so it doesn’t pollute my browsing history) and DuckDuckGo shows the stupid AI overview because of course, in incognito it’s not gonna know that I specifically disabled their “AI” features. It should be opt-in!!!

Yes, they’re kinda useful by themselves, but in a search engine? Sure, it can filter results or whatever, but it’s just easier for me to go to a website I know and look for the info there.

Also, I don’t like chatbots because usually I just want to search one thing and leave - and to me, it doesn’t feel like that’s what they were designed for. They’re meant for “conversation” but I don’t want to talk to a bot. There are, of course, a plethora of other reasons why I don’t like LLMs, which I won’t be getting into in this reply.