Skip to main content


How often do you use ChatGPT?

#EvanPoll #Poll

  • Daily or more (7%, 37 votes)
  • About weekly (15%, 78 votes)
  • About monthly (13%, 68 votes)
  • Quarterly or less (64%, 329 votes)
512 voters. Poll end: 1 year ago

Evan Prodromou reshared this.

Hmm, I think there should be an "Avoid it on principle" option?
@timbray that's "or less". See #EvanPollFAQNever .

What's the principle?
@timbray

"Avoid feeding the mouth that will soon bite you"
I said about once a week, with the understanding that “using” ChatGPT includes fucking around with ChatGPT.
@BathysphereHat that's fair! I've used it for some research for work, for recipes, and for gardening, but most of the rest is just fucking around for me, too.
If you want to say "never," reply "quarterly or less". See #EvanPollFAQNever .
I've done so, but still wanted to explicitly pronounce my distaste for the technology and for the marketing hype and fanboi fawning. Deep breath... and move on 😀
@stevefenton what don't you like about the technology?
This is a rare case where the technology just cannot be viewed as neutral. All of the "it will save time" labels have missed just how much time will be wasted by the information pollution. This is the Deepwater Horizon oil spill of The Web.

I have a longer answer here...

https://www.stevefenton.co.uk/blog/2023/02/facing-reality-ai/
Not good enough, sorry. This is one poll I will vocally skip.
nah, that'll skew results and big up hype for it, I'll just wait for ya to release the tally
Never is qualitatively and quantitatively very very different from "quarterly or less"
@baldur I realize it's very important to some people, but it's not interesting enough to me to have it take up a whole option.
Swap word "train" for "use"

and you'll be on to #Musk + #Thiel strategy in that poll

🧐

#OpenAI
#ChatGPT
@TryshHQ is that true? Does the data from my sessions inform the next generation of the model?
Following #MentalHealth #chatbots since 2017 (all of them, most egregious WoeBot at Stanford).

*ANY* chatbots main job is to harvest data & behaviours from humans using them to improve all aspects of their own back end systems.

#OpenAI ToS/EULA “OK” = Liability to User

115 days and #ChatGPT will be in MS-Office within weeks?

WTF?

#AI = savage & lawless land
#VC’s❤️

ALL #media dopamine addled or captured or troll army pumping #AI for #Musk #Theil

Who’s looking behind curtain? 🧐
absolutely: https://help.openai.com/en/articles/6783457-chatgpt-general-faq
  • Will you use my conversations for training? Yes. Your conversations may be reviewed by our AI trainers to improve our systems
So this is the thing... I actually *want* to use ChatGPT more. I think it's super powerful and can do a lot. The thing is, I'm not finding moments where I actually need it. It's faster to write code myself and most stuff I can find on Google faster than coming up with a prompt. I don't want to get left behind in the next wave of tech evolution, but I'm struggling to find ways to use it in my daily life (even outside of work).
@matthewjones So, I use it for cooking, for cocktail recipes, for code, for research for work, and for gardening.
Wow! Maybe I just need to push myself to use it more. Perhaps it's more me just breaking my habits. 😀
I haven't touched it, and I doubt I will ever use AI tools as tools on a regular basis. It doesn't fit my personal amity.
Maybe it is useful but I don't even read articles about ChatGPT. ¯\_(ツ)_/¯
Fucking never. Ask me again after the fad wears off. Same answer
@evan
my poll answer: Not at all … yet.
I have used it twice; I think that LLM can be interesting as a sounding board in areas where I am knowledgeable enough to understand falsehoods and hallucinations in the response. However, the current lack of transparency regarding the training architecture and data sets, as well as third-party revelations regarding training have me ethically opposed to the current tools and companies deploying LLM. There is potential there when coupled with other models and with better transparency
I don't know what that is. Not yet motivated enough to find out either. I've seen people mentioning it on various social media platforms.
Never? Tried it a few times, but beyond that..
@evan
As in, I tried it when it was new, and never looked at it since.
Never. At least that I know about.
This is a really interesting poll.

I'm an about-weekly. I use it for cocktail recipes, cooking, gardening, and research for work and school.

I've found it really helpful. I almost never use the output directly; it's almost always for research.

Occasionally I'll share the content directly with colleagues; "here's what chatGPT says." I haven't ever passed off the output as my own work.
I think for me it replaces my previous research process: complicated Google queries, reading through Wikipedia articles and websites, and then writing it up as bullet points.

I know a lot of people are upset that chatGPT doesn't credit original authors, but I wasn't doing that in my previous work for private research anyway. 🤷🏽‍♂️
People will present ChatGPT as their own and are trying to train people that whatever it spits out is correct. It’s depressing
i mean, you might not have been citing them, but you were probably evaluating their trustworthiness on some level subconsciously, right? like you're not going to trust a fox news or breitbart article
@shoofle yes, and I'm doing that with ChatGPT, too. I use it as a jumping off point for research, and don't take it's output as gospel.
well yeah, but you can't examine chatgpt's sources.
@shoofle sure, but I don't need to ensure that its recommendations for berry bushes in southern Quebec's climate come from unimpeachable sources. I just look up those species on Wikipedia to double-check.
I use it for ideas for titles of papers. I provide a summary and the audience (usually security, usually #DefCon), and I get a few kinda funny choices to riff on.
the cocktail recipes it gives are pretty amazing sounding! And to be be fair I give it “themes” and it provides good ones!
updated reply from yesterday’s “not at all” today i got an openai api key
an experiment on digesting a large wiki’s content and requesting analysis by topic or theme or … i am not sure what i am looking for or what i may find or notice