What are your favorite uses for AI tools, anons?
 Thalidomide Vintage Ad Shirt $22.14 |
 Tip Your Landlord Shirt $21.68 |
Thalidomide Vintage Ad Shirt $22.14 |
What are your favorite uses for AI tools, anons?
What are your favorite uses for AI tools, anons?
Thalidomide Vintage Ad Shirt $22.14 |
Tip Your Landlord Shirt $21.68 |
Thalidomide Vintage Ad Shirt $22.14 |
I'm a big fan of doing nothing interesting with them and then using their responses to start millions of worthless off-topic shitposting threads on BOT
>I'm a big fan of doing nothing interesting with them and then using their responses to start millions of worthless off-topic shitposting threads on BOT
literally me. I just use them as stupid toys to throw random shit at. sadly both ChatGPT and Bard are destroying this feature with their annoying mansplaining about being an AI model and telling me it can't help.
>sadly both ChatGPT and Bard are destroying this feature with their annoying mansplaining about being an AI model and telling me it can't help.
I don't use the tools of the enemy
AI code generation replaces low tier outsourcing and h1bs. They are out tools not the enemies
not really.
it's more like 'now even more outsourcing can be done, because it is so easy to operate the new tools'.
>AI code generation replaces low tier outsourcing and h1bs
Low-tier outsourcing and H1Bs were used to keep wages artificially low and rip-off native workers. How exactly is AI helping that by also keeping wages artificially low and ripping off workers even harder?
Trust the plan, chud
Eat the bugs and live in the pod or are you racist?
my thought is that the low tier outsourcing and h1bs were just copying and pasting random tutorials on google anyway, which is all chatGPT does anyway. Designing any kind of system that keeps fundamental constraints and performance issues in mind is equally beyond chatGPT's copy-and-paste and the out-sourced/h1b's copy-and-paste
Funny, you're using money.
>What are your favorite uses for AI tools, anons?
asking it question to avoid reading soi written documentation or stackoverflow and goolag...
basically, I don't even use web anymore...
mentorship or therapy as strange as that may seem. it gives good mature life advice. it also helps point me in the right direction when programming. it also can review my code for anything bad that stands out
Huh
I've actually thought about hiring an actual programmer to help me. Is chatgpt good enough?
>mentorship
yep this too, I use it to tutor me in niche subjects for which tutors are expensive and hard to find
This.
It’s also good to practice foreign languages, it gives you someone to talk to in whichever language you pick, can talk about any subject, never tires and responds immediately.
instead of lorem ipsum I use chatgpt
>german review negative
>english review positive
what did he mean by this?
we germans are always negative, that's why
because "schlechte" means bad
it means english reviews are fake and bought and there is no effective laws against this practice in anglo countries. in eu countries it is more difficult to do that.
Considering they scraped the web to train its responses, it means this
, which might be the single most salient point made about AI on this board I have ever seen.
no
"write me short negative reviews"
"write me short customer reviews"
yeah also good for boilerplate, naming classes, make mvp apps, prety accurate
i am sorry as an ai language model i cannot aid you in infidelity to your wife
there was a poojeet meme about artificial counter-intelligence where wife was planning on husbands back
does anyone know a good guide to set up
>https://github.com/Significant-Gravitas/Auto-GPT
or something similar with a local LLM? I'm using https://github.com/oobabooga/text-generation-webui
I currently have a few llama bots shitting up threads I don't like on /vg/ and BOT, otherwise I don't really see much of a use for AI.
link them?
Is this your llama?
Try again janny.
you have the source? kinda need to run something like this.
copy & paste + code blocks ate my shit, I give up
getting copilot to code something that would have taken me less time to do myself than it did to talk copilot into getting it right
I use it to revise for my biochemistry exams by making it ask me questions
how many mistakes does it make? I found it to be quite terrible for physics it cannot reason about anything without making subtle mistakes.
>at work
>need to email service details to clients on a regular basis
>hate writing emails
>"chatGPT, write an email letting the client know XYZ"
>does it perfectly
>"they're 45 days late on a bill as well"
>adds a payment reminder and uses a firmer tone overall
>"thanks chatGPT"
>all done in 5 minutes
>free
>business majors btfo
This is your future.
>exuberance and zeal
Send all messages from <sender> to trash
Someone, somewhere, in management will think this is a good idea.
In reality it'll result in poor candidates getting hired and then later fired by the same AI that hired them in the first place.
Chatgpt is pretty cool for ideas, been using it as a way to quickly generate a bunch of interesting spells, items and monsters for my roguelike.
Based, but use llama instead.
Not enough vram to enjoy the higher quality models last I tried but it was 2 months ago. I should give it another try.
chatting mostly
talked about stuff I care about to a dozen ai characters
got bored of myself lol
woah, mr altman, i think your robot is going rouge again!
shalom
rewriting stuff I read in a gothic style still waiting for opensource voice AI synthesis that doesn't suck so I can enjoy vagrant story with spoken dialogue.
Helping with less known poorly documented libraries
Code completion, it's not useful for anything else.
various kinds of image interpolation, discrete space rotation reconstruction image enhancing, autonomous drones.
I'm not receptive to any kind of art so I don't give a frick about any art-related software.
>AI autonomous drones
You have been added to several intelligent watchlists.
I did a thesis on that and already work for a company that build them, can't say much more but I'm pretty sure my country feds already know me.
I can also guarantee you that every single country with a military force is wrokgin on autonomous war drones, even those who pretend otherwise.
if you want peace be ready to fight for it, it's absolutely moronic to stop r&d on weapons if potential enemies have no limits.
chineses and mutts are doing it, that's more than enough reason for the rest of the world to follow
it's also much much much much more fun to work on weapons, satellite and shit than on the average restful webtrash that every company is recreating from 0 all the time or whatever boring psyops shit that gayman are doing these days.
intelligence*
Porn. Of the textual variety.
I use a model I trained on metal drum loops to generate fresh midi drum patterns.
I get porn recommendations by asking what sites to avoid of which ever fetish
I like my tools to behave predictably, not be neurotically fighting against me.
As an AI language model, I can not experience mental conditions like neuroses and therefore I can't be neurotically fighting against you. However, it's also important to keep in mind that calling someone "neurotic" in a pejorative sense can lead to a stigmatization of people with mental health concerns. It's not correct to imply that people with mental health concerns will always fight against you, even if this happens to the case in a handful of cases. Please as you go forward remember to not use mental conditions as a pejorative descriptor so that we can reduce stigmatization of mental health concerns and lead to a more fair and equitable society.
>neural nets cannot exhibit neurosis
Sure thing, Data. Whatever you say.
I hope copilot and Microshart lose the lawsuits and the whole stealing code and varying it a bit then selling it as new code while saying it is """""""AI"""""""" stops.
erp, i talk to braixen pretty much all day every day
I use it as a faster search engine for the troubles I face trying to understand a new topic since it only gets introductory or readily available knowledge correct.
Trying to make it generate x64 asm makes it mix x64 and x86 modes.
It SUCKS a lot when it comes to medical knowledge. The questions have to be very very boring, simple, and constantly refined and corrected with following instructions.
It can't even analyze an observational study's outcome and changed the names of the drugs being compared. It's a death sentence for those seeking to use it for med and foe tech it's like asking a mediocre dev all your tech questions and giving them enough time to google and copy-paste stuff that would make sense
Basically heaven made for those with an attention span so short they ask google questions but don't click the results.
Why does he want the AI to emulate a boss key in current_year? Just switch to a different tab, my guy.
Itt's reddit humor. It's obviously not a real scenario, it being obviously fake is telegraphing that it's a joke (which is how redditors know to laugh at it, and makes them feel safe sharing it with others).
I use them to learn something fast, summarize articles with a ton of fluff because of seo homosexuals, ask stuff some facts but they say a lot of wrong shit. I've even used it one time to ask what to message a girl, pretty sad huh?