please vote anonymously and boost this poll for transparency and participation:

[falcennial@mastodon.social]

@thejessiekirk oh yeah. the full blown AI psychosis cases from doing so are intense.

although I'm sure there are many more non-harmful instances of doing so. it being a normal private activity that the news couldnt and wouldnt report on the way they report on the AI killings and so forth.

but yeah, people do.

[janxdevil@sfba.social]

@falcennial So as it happens, I am the On-call Engineer for the core platform team this weekend. The first and only thing I have been trained to do when the pager rings is to log into the system and run a custom shoggoth skill to start responding to the incident. What am I supposed to do if the shoggoth is having an outage? Wing it, that’s pretty much the plan.

[abramkedge@beige.party]

@Mekachu04 @falcennial Oh... I had forgotten about that. I don't think it counts, because all you're trying to do is get to a human in as few keystrokes as possible.

[johan@s.cafe]

@falcennial

The option "Several times a day, but less than an hour" is missing.

[theriac@plasmatrap.com]

@indigoparadox@mastodon.social @falcennial@mastodon.social
startpage used to be amazing a few years ago, then they decided to switch to primarily using Bing for some reason.

Worth pointing out startpage added an AI option a few months ago. They say it's optional but I haven't seen any assurances that your actions and/or searches aren't training their pet AI.

[annabanana@rivals.space]

@falcennial i use only text LLM to help me with computer science advice, linux (i double check the command it gives). The thing is i struggle a lot even though i'm on linux for 10 years, and i have often weird issue i struggle to find on the net. Feel ashamed at first but hey here we are

Using it sometimes made me open a little the domain

I may be caught using it for anxiety. For example if i have an important thing to buy and i don't want to take the wrong thing. It can help me to calm down and make the right choice. The thing is i always have the feeling that instead i could use social media, but yeah my brain kinda like not to bother people

I don't say it is justified or good. I'm not feeling great telling this. I may reassure myself saying "it's just text, no image analysing etc". But yeah here i am. I try to reduce, i see how it can be attractive

[gilleslecorre2@toot.aquilenet.fr]

@falcennial
I even don't know what it is.

[aedius@lavraievie.social]

@falcennial @jaseg

Thanks I had the same question.

I never do casual chat with it, I only give it ticket at work.
And when I see how random are the response ( sometimes good, many times very bad ) I don't want to chat with this stuff.

[missbullitt@masto.bike]

@falcennial do you include discussions with companies support chatbots?

[bobthomson70@mastodon.social]

@falcennial I use such tech mostly for work but I do not understand folk ‘talking’ to these things. I input questions, like I used to do with Google when it wasn’t full of AI slop and enshittified, but I do not engage in dialogue with an automated sentence guesser and search engine combo.

[rinaldogiorgetti@puntarella.party]

@falcennial what's the meaning of "chat" in this case?
- Ask technical queston
- General conversation
- both

[noisecolor@toot.community]

@falcennial
I do it probably a few hours per day during work for images, copy, coding, general questions,.. and then a few times per day at home, depending what I'm doing.
When I tinker with home server I can do it for hours at home too. And sometimes when I'm doing or wondering about other stuff or just playing around.

[jigmedatse@social.openpsychology.net]

@falcennial@mastodon.social I get bored, but then I realise that LLMs are even more boring than what I'm bored of. I haven't learnt to just not bother.

[noisecolor@toot.community]

@failedLyndonLaRouchite
It's way better: no toxic people, civilized conversation and generally accurate information about most non-specialized stuff.
No chance someone will wish you death because of some banal opinion you might have, no trolling.

LLMs are way better than people in that regard.
@falcennial

[thejessiekirk@ohai.social]

@falcennial I'm not so sure. Chatting implies an informal conversation between two or more people, so I think people *believe* they are chatting but aren't really.

[gregalotl@c.im]

@falcennial
As a non-coder at a well known large software developer, I have a backlog of ideas I want to realise in code. Until AI came along, I had no way to turn aspiration into reality. Now I do! Feedback has been that my ideas are useful (ie not self delusional ideas without actual merit)

AI for research summaries useful, and removes the drudgery but should be checked + explicit instructions to explain assumptions & extrapolations.

Generative AI, chatbots no use to me. Don't see the point.

[emily@bark.lgbt]

@falcennial wait, chat or use?

[emily@bark.lgbt]

@falcennial I have made a mistake with my answer

[derpumu@mstdn.games]

@toolsontech @falcennial yup. "Make an analysis of a couple hundred files of source code, find dependency cycles and tell me the most effective way to break them" - it chews on the stuff for an hour while I have lunch, and then spits out a concise document what the best course of action is (or close enough to best, no way to be sure, but 90% is good enough). Instead of me having to pour over code for days before I even make a dent.
But having a chat for entertainment? Nah. Not worth the money, energy spent, etc to get some regurgitated bs the bot found on the webs

[agowa338@chaos.social]

@falcennial Ngl, I'd prefer not to, but at this point it is seriously way better than google search...

(And that is mostly because of how enshittified the google search has become. Maybe I should just take some time and setup YaCy...)