Thread 7614878

aside from the slopshit, AI is suprisingly useful to design improved curriculums for various books, programs and such that were vague.

what prompts should i use to make AI teach me drawing better?

>subhuman averse to reading a little bit
Holy shit kill yourself and delete your thread.

>>7614880
>cryptid ass pseudo texts with useless scienceslop bloats
dont blame me for using AI to give me what the fuck are they actually saying.

>>7614878 (OP)
You have to actually know something about the topic in order to be able to tell if this is working. For education it's more for a teacher to build lessons than autodidacts, because you lack the foundational knowledge to check if it's good or not.

I suppose it's not much worse than asking some random artist (who could be a terrible teacher or a crab) or getting your art lessons from twitter, though. Some of the advice on this very board has been so rancid an AI would at least let you gamble for a better response faster.

>>7614882
Are you literally unable to think? AI destroyed your brain it seems.

>>7614878 (OP)
https://arxiv.org/abs/2506.08872
>using ChatGPT atrophies the parts in your brain which are needed to learn things
bummer

>>7614878 (OP)
Is this a joke thread. The book is already a curriculum with checkpoints for you to evaluate your progress. Why would you need this?

>>7614878 (OP)
>permabeg trying to slack off and skip the work
And that's why you're a permabeg

ITT: luddite seethe

>>7614897
>https://arxiv.org/abs/2506.08872
>We assigned participants to three groups: LLM group, Search Engine group, Brain-only group, where each participant used a designated tool (or no tool in the latter) to write an essay. We conducted 3 sessions with the same group assignment for each participant. In the 4th session we asked LLM group participants to use no tools (we refer to them as LLM-to-Brain), and the Brain-only group participants were asked to use LLM (Brain-to-LLM). We recruited a total of 54 participants for Sessions 1, 2, 3, and 18 participants among them completed session 4.
Doesn't say what you think it does. Sorry you lost your Polygon job!

>>7614909
Using ChatGPT makes you smarter?

>>7614913
Never said that.
What it does do is let smart people get more work done faster, reducing the cognitive work needed to perform more-menial tasks so that more-advanced tasks can be done sooner (not faster - sooner). Picrel is an example.
It also filters retards by telling them to eat glue, and filters midwits by the bad info being the equivalent of wide-scale adaptable breakfast questions.
You know you're talking to someone who only knows how to SOUND smart if they think "13 digit multiplication" means 13x13.

>>7614913
Oh also one addition >>7614918
That's not using chatGPT, specifically. Just AI tools in general. You need to use the right tool for the job, and every LLM is going to be different.

>>7614918
>Never said that.
Does it make you smarter or dumber then?

>>7614918
>It also filters retards
Sounds like cope for using an absolute shit tool.
>reducing the cognitive work
And makes you retarded yeah

>>7614923
Humans are cognitive misers. We only put as much effort into things as is needed, because brains consume a shitload of calories and generate a ton of heat. Moving cognitive drudgery from being in your head to being performed by silicon is a gain in time and effort. And this also includes tasks that are a mix of physical and cognitive work. Doesn't matter if it's AI, word processing, physics simulations, electronically-controlled security systems, home printers or digital art filters.

>>7614922
It doesn't "make you" anything. Books don't make you smarter either. To be smarter you have to work to eliminate contradictions and improve your thought processes.
That comes from just intaking the right information, which can come from anywhere. Media (including books, video, etc), random generators, other people, observing nature. That's a part of why metaphorically "touching grass" is important. If you limit yourself to learning from a single environment then the scope of what you randomly encounter (and thus can learn and grow from) is also limited.

In a perfect world, AI would empower people to optimally develop their skills in a way specific to them rather than replace them or the need for human input. I don't see that happening in this world though.

>>7614878 (OP)
I use stable defusion and some LORAs to use as figure drawing reference practice. It's worth it if you have the patience to dial in the proportions and poses you want to study.

>7614941
I swear is this the tripfag from last year for amount of shilling this nigga makes for it with pseud logic
I like how you talk about "intaking" the right information when AI has made shit up on several occasions but will prob still go "I'll ignore"
>you'll forever be a pajeet anon

>>7614878 (OP)
Personally I think you should just read the book - it tell you about a technique, and then gives you an assignment to practice said technique (if I recall keys to drawing correctly).

Other books, like say Loomis' work, also have assignments, but there are like 2 or 3 for everything he was taught, and none are specific to any one skill - that'd be a good book to ask AI to create some assignments for you, per lesson/page/chapter in the book.

Loomis and other drawing book authors can sometimes also be a bit dated or flowery in their language, and it can make understanding exactly what they mean somewhat confusing, so I suppose asking AI to explain their lessons, or give you bullet points, or whatever, would help as well.

Finally, if there's a particular drawing assignment/task you're having trouble with, you could ask the AI to give less difficult assignments to you that ramp up to the one you found challenging, that all cover the same concept.

There you go, some shitty ideas for you, but personally I think you should just read the books and stop having AI give you plot summaries of your adhd zoomie zoomer tik-tok videos, it can't be good for you.

>>7614971
Information does not have to be REAL WORLD FACTS ABOUT BUGS to trigger an improvement in your thought process.
I sometimes use a LLM to assist with worldbuilding. I end up typing paragraphs and paragraphs and editing and revising. And a lot of the time it gives me boring, inane responses and tries to orbit back towards mainstream shit. Sometimes a single phrase or sentence helps spark a whole set of ideas though. That's really the point. I've had more "Wait but what if..." moments since I've been using these things than before it.

But also, if you are the kind of person who is used to questioning anything someone says, then using an LLM to explore ideas, like a conversation partner, can be useful. However! If you are the kind of oversocialized brainlet who is used to simply trusting what other people say, uncritically, because that's what's "normal" for you, then using AI will give you brain rot, sure. Fuck that practice though. People should not be doing that and I encourage the destruction of retarded cultural practices.

>>7614883
What kind of advice or "conventional wisdom" do you see taught to begs on here that you consider to be rancid?

>>7614997
>thread about learning a skill of visually depicting events, beings and objects
>heh well actually sometimes I ask chatgpt to make shit up and larp as my dommy mommy goth succubus gf, it triggers an improvement in my thought process

>>7615000
A lot of it involves things like grinding for hours at a time, telling people to draw FOTM shit, or "advice" on how to use twitter that will put you in algorithm hell. Someone in another thread even wrote that you should draw using your wrist.
There's a lot of really terrible info on here that is impossible to accurately discern as to whether it's intentional crab trolling or blind retards leading the blind.

>>7615006
If that's what you took from what I said then you are ngmi as your low IQ imagination is stunted without even needing to burn out on deepthroating AI generated pizza recipes. 15 years ago you would have used the crystal recipe someone posted on /b/ and gassed yourself.

>>7615012
>your imagination is stunted
>computer, what do the orcs eat? provide 5 answers so I may edit and revise
lol

@grok is this true?

>>7615025
More like, I paste in the entire writeup I have for multiple things in a single thread and it gives feedback, comments, suggestions, etc.
I then critique its responses and clarify some things. Helps with making the thematics and mood I am going for more clear and leads onto other threads of reasoning.

If you are asking an LLM simple questions rather than typing paragraphs upon paragraphs of context and other things, you are going to get generic responses.

The alternative is to find a similar individual with the time and energy to devote to talking about something in detail, and is willing to return to the same topic over and over again at random points over the course of weeks, while maintaining memory of the conversational context.

>>7615038
holy retard also nobody asked

kek did the tripfag find out about the disney lawsuit?
he always goes into shill mode when something bad happens

>>7615039
Don't you have some overpriced pokemon cards to buy off ebay before bitching on reddit about how much money you gave a scalper?

lel the bot broke

>>7615052
>No no no I'm not coping! Real people don't exist who disagree with me! It's all a conspiracy!! Shills!! It's bots! It's Russians! It's street shitters! It's FBI CIA Glownigger Kikes!
This is what you look like to me.

>>7615056
it's probably time for your meds

>Shills!! It's bots! It's Russians! It's street shitters! It's FBI CIA Glownigger Kikes!
the bot spilled the beans, better fix it, elon