Embarrassing

>implying other ai's can do better

>>105612226 (OP)
It is because musky injected South Africa white genocide in the prompt (he is afraid he will lose his colony)

>>105612238
Gemini. None of the AI can draw

>>105612265
>>105612226 (OP)
I dont understand how they're all competitors with their own chat bots but they all get it wrong in the same way. I've noticed this with other things in programming also.

Is there some single entity training all this shit and selling them the models?

>>105612226 (OP)

>>105612321
Yes, look up Scale.ai. They all buy training data from the same data farms.

>>105612265
Hey, we're getting closer bros

can't wait for ai to steal our jobs so we get ubi

>>105612291
I swear it's the only useful one.

>>105612481
>time wasting question

whenever i read "ai can't do X" i always think of how well this meme aged in 1 year

>>105612321
>single entity
>all this shit
rly makes u think

>>105612226 (OP)
you didn't say draw it to scale

>>105612238

>>105613451

>>105612481
You mean jewish kill shot vaxxes?

this will be stRawbeRRied into the next iteration and we can all rest easy

>>105612321
Well, it's all roughly the same data, the same technology, and the same bunch of greedy retards developing it.

>>105614840
i’m jewish tho

>>105614362
Cope, AI is the future, humans will soon be obsolete.

Microsoft wins again ;^)
main :: IO ()
main = mapM_ putStrLn (triangle 10 10 2)

triangle :: Int -> Int -> Int -> [String]
triangle base height thickness =
[ row y | y <- [0 .. height - 1] ]
where
row y
| y < thickness = replicate (base - 1 - y) ' ' ++ replicate (2 * y + 1) '*'
| y < height - thickness = replicate (base - 1 - y) ' ' ++ "*" ++ replicate (2 * y - 1) ' ' ++ "*"
| otherwise = replicate (2 * base - 1) '*'

>>105615145
>I can't draw a triangle, but I can generate some code for you to upload to an image generating program!

>>105615074
cope? did you even try to contextualize what i said?

>>105612238
Wolfram Alpha can, but that's cheating.

>>105615379
It took a bit of digging to get to it, but it is on point.

>>105612265
I think this is solvable.

yeah ok

>Let point A be at [0,0]. Let point B be at [10,0]. Let circle C be centered on B, and be radius=2. Let point D intersect circle C, and let the length AD = 10. D is in the first Cartesian quadrant. After solving the coordinates of D, create a simple 2D dxf file with these lines: AB, BD, DA.

Model: Qwen/Qwen3-235B-A22B

If you show thinking it sounds lunatic. But it does output a dxf that imports perfectly to Blender, as you can see.

>>105612226 (OP)
>>105612265
>>105612374
Try again in 100 years, when AI has got better and of any use whatsoever.

>>105615695
anon... I am from the FUTURE, with a picture of the magic triangle

>>105615145
You should ask for a 2d dxf

see
>>105615660

>>105615695
Nah, just wait 1-2 more years.

>>105615863
or negative 1 hour, naomi

>>105612321
LLMs don't see images, at best they tokenize images and then detokenize them, which somewhat works but utterly breaks when precision is required.
This would require VLMs but they are much more compute and memory intensive. Another approach would be a general diffusion model that's trained on text input as images (in addition to images).

tl;dr it's a resource constraint