Will Clark
josefi think the best way to use LLMs is this: don't type into the AI chatbot's input field. instead, write a description of the problem in a text file. work through the specifics of what you want to accomplish, and how you'd go about doing it. once you've got a few hundred words of ideas and planning, you can go back to the empty, unused AI tab and close it. then begin doing the work yourself. consider sending your notes to friends or experts. when you're done, remember to thank them for their help!
Random Geek@jk I'm saving that one.
@ NovaNaturalist🇨🇦🇩🇰🇬🇱🇵🇦🇲🇽🇱🇸🇳🇫🇺🇦🏳️🌈 #FBPEI also find the best way to use AI is to entirely ignore an algorithm which is merely designed to generate something that looks plausible - and then to use your own reasoning and logic.
Jigso@jk the best AI for debugging software is the rubber duck (https://en.m.wikipedia.org/wiki/Rubber_duck_debugging) but I am pretty sure it works in many domains
🌸 Vårgubben@jigso @jk As I’ve come to understand it, it’s only considered ”AI” if it’s mostly useless. We don’t consider the statistical models in auto complete (or spell check) as ”AI” because it kind of works fine. Same with facial recognition in photo album software, or navigation in robotic vacuum cleaners.
Thus, since the rubber duck works it can not be considered fulfilling this criteria for ”AI”.
Andreas K@jordgubben
But it is not useless.
You can use LLM for useful stuff.
But telling a LLM to write you a pink pony does not work. (Literally from a company internal presentation.)
Breaking the job down into multiple steps (doing the architecture by hand), and then using a LLM to process the text. And extract data from text. And that worked surprisingly well.
Now most AI assistants for coding are rather primitive on the algorithm side how they deal with the LLM.
Simple example
@jigso @jk
wb x64
Flatbush Gardener 🌈
Octorine
Nick@jk I've actually found myself often typing in a prompt for a LLM and while trying to figure out how to describe my problem to the dumb machine, I found a new idea on how to solve just that problem. So yeah many AI prompts can probably be replaced by an text editor
Francis (is looking for work)@jk so a self-hosted LLM, but the host is you and the self is also you and the LLM is also you. Got it.
refraction 
@teacherbuknoy @jk the holy trinity
UsagiJer 🐰 ಜೆರೆಮಿ@jk Artificial Rubber Duck . call it ARDaaS , and sell subscriptions
Clinton Anderson SwordForHireBut... But... But.... If I don't use "AI" how will I ever demolish the environment while churning out mountains of garbage based on stolen content and forgetting how to actually do anything?!?!?!
"I used AI to....", is nothing more than, "Listen I'm not an asshole but....", for the 21st Century.
FoolishOwl@jk I've wondered if the people who are required to use LLMs in their jobs could basically follow this method while piping LLM output into /dev/null.
Magnus Ahltorp
Martin Vermeer FCD@jk Like using aspirin as a contraceptive... between the knees
chesheer@jk We need to wrap it in a more flashy terminology for Gen-Z.
We need to give this a name... like SelfGPT?
"Did you know there's a hidden subsystem in your own mind that can work like ChatGPT, but entirely eco-friendly and FREE? Here's a secret method for unleashing a true potential of your SelfGPT..."
Then a hundred more TikToks about this secret technique, and we can call it a day.
We need to give this a name... like SelfGPT?
"Did you know there's a hidden subsystem in your own mind that can work like ChatGPT, but entirely eco-friendly and FREE? Here's a secret method for unleashing a true potential of your SelfGPT..."
Then a hundred more TikToks about this secret technique, and we can call it a day.
Morgan ⚧️
Alex Keane@raphaelmorgan @chesheer @jk Like millenials joking about the forbidden candy of tide pods (mostly) didn't eat them, or a lot of Gen X had parents who did remember they had kids?
Wild. Can't be true.
(To be clear, the humor is meant as agreement)
Zeborah
Felix
PurpleShonaS@jk masterful!
Flaming Cheeto@jk ok, but I'm using the LLM to send the thank you notes
ZoLo 🏴🇳🇴🏳️🌈🏳️⚧️♾️@jk
The amount of times I've figured out the solution to my problem in the process of writing out my question on Stack Exchange, so I didn't even have to post it 😂
The amount of times I've figured out the solution to my problem in the process of writing out my question on Stack Exchange, so I didn't even have to post it 😂
Vadim Rutkovsky
Willow "Wolveric" Catkin
properlypurple
Jay Hannah@vadim @jk @genehack lol @eric taught me to use 🦆 emoji in Slack to indicate when I was https://en.wikipedia.org/wiki/Rubber_duck_debugging
@jk thanks
Stuart Longland (VK4MSL)@jk I use the Jira comment field on the task to do this.
Not using their AI, just leaving a comment with my thoughts at the time. It's then a historical record not just for me, but also for the reviewer.
AnneH@jk I may or may not regularly walk up to my coworkers and ask if I can use them as a sounding board literally so that I can hear myself explain a problem to someone else.
(The bonus when talking to a coworker is that sometimes they *also* get an idea and then I have TWO ideas to go back with)
Pete@jk I use LLM as a Mechanical Turk.
@iveyline@jk A perfect solution. Use real intelligence!
ENIGMATICO 
gay jay@jk better yet, just post all of it on fedi as you're thinking of it, and sometimes you can pull a real person along for the ride
Pavel Machek@jk Yeah, your friends and experts will be happy to do the work chatbot, google search, or stack overflow search could do.
¯\_(ツ)_/¯
➴➴➴Æ🜔Ɲ.Ƈꭚ⍴𝔥єɼ👩🏻💻@jk I'm incredibly amazed at how many Anti-AI people just assume thaat everyone has the same cognitive and social abilities and advantages that they have.
Rog Dolos 🏳️⚧️@jk LLMs. It really is just rubber-ducking.
Nick Radcliffe@jk My friend Patrick had a related idea for playing the lottery. Each week, you pick your numbers but don’t buy a ticket. Then you watch the results come in and enjoy not having wasted money on a losing ticket. You save a small fortune.
Of course (in the case of the UK lottery) 1 in 14 million times you will feed sad* when your numbers come up, but what are the odds?
* I think Patrick actually said “have to kill yourself”, but that seems a bit extreme.
Chris King@jk absolute banger post 😘 exactly my thoughts on the matter
Dr Susi Arnott@jk Thanks a million; and hadn't come across this term before :) #rubberducking
https://en.wikipedia.org/wiki/Rubber_duck_debugging
https://en.wikipedia.org/wiki/Rubber_duck_debugging
V
Coach Pāṇini ®I don’t bestow this score lightly: 17/10 for trollsmanship.
Exquisite #shitposting form.
Anna Liberty@jk This is why I love literate programming
synlogic4242@jk upskills one rather than deskills. deskilling is a trap that capitalist psychopaths want to lure labor into
sammypanda@jk i love experts and know very few 😓
Jeff Atwood@jk you mean .. like .. talking to a rubber duck? https://blog.codinghorror.com/rubber-duck-problem-solving/
Rubber Duck Problem Solving
At Stack Exchange, we insist that people who ask questions put some effort into their question, and we’re kind of strict about it. That is, when you set out to ask a question, you should… * Describe what’s happening in sufficient detail that we can follow along. Provide the