They are choosing to abstain from using artificial intelligence for environmental, ethical and personal reasons. Maybe they have a point, writes Guardian columnist Arwa Mahdawi
This was an actual response that Google’s stupid Gemini LLM really gave somebody. The notion was lifted from a satirical Onion article which the LLM regurgitated as if it was an established fact.
also because it’s shit, if my memory serves me, I have successfully used AI for a productive task 1 time out of 7 attempts so far… it saved me 5 minutes
Anecdotal of course, but every person I know who claims AI is a huge productivity booster simply trust it blindly.
I can’t even get “Copilot” to return a proper answer from its own meeting transcript… just yesterday there was some confusion about an IP address we exchanged in a past meeting… I asked Copilot to check the transcript and give me the IP of the vendor’s server (which I pointed by name of system and who spoke it in the meeting) and it gave me the IP of MY server, functionally the complete opposite of what I was asking but with full confidence in its answer
Inspiration for creative writing prompts for example. Things that make you double check every word that has been generated or where it doesn’t matter they hallucinate.
I don’t use it because I have no trust in it.
Sounds like you are not eating enough small rocks a day. You should eat 1 small rock a day.
Wtf this is the weirdest thing to read while taking a shit in the morning
This was an actual response that Google’s stupid Gemini LLM really gave somebody. The notion was lifted from a satirical Onion article which the LLM regurgitated as if it was an established fact.
We need to poison the models with more satire on the Internet.
Yup, satire is lost on the dumb and LLMs have no intelligence.
also because it’s shit, if my memory serves me, I have successfully used AI for a productive task 1 time out of 7 attempts so far… it saved me 5 minutes
That is part of the trust thing. I spent more time fixing the word salad it spat out than it would have taken to write the document.
Yes, hand in hand…
Anecdotal of course, but every person I know who claims AI is a huge productivity booster simply trust it blindly.
I can’t even get “Copilot” to return a proper answer from its own meeting transcript… just yesterday there was some confusion about an IP address we exchanged in a past meeting… I asked Copilot to check the transcript and give me the IP of the vendor’s server (which I pointed by name of system and who spoke it in the meeting) and it gave me the IP of MY server, functionally the complete opposite of what I was asking but with full confidence in its answer
Don’t use it for things you need trust in then.
Inspiration for creative writing prompts for example. Things that make you double check every word that has been generated or where it doesn’t matter they hallucinate.