"Drawing" the Mona Lisa with GPT-5.6, Claude, Gemini, and Grok

Posted by hershyb_ 3 days ago

Comments

Comment by NichoPaolucci 3 days ago

As I looked through the images I was unimpressed entirely, at first. But, then I started thinking, these look a little... "childish" to me.

Childish as in... A newish artist who is drawing a concept rather than light / forms (Which is something artists typically do as they understand drawing more and more).

The rose in the vase specifically - some models understood that there was supposed to be shading, reflections, the concept of refraction - others just drew "blue = glass" and "green = stem" and "red = rose".

Really odd to look at, considering if I saw any of these drawings from a human kid, I would say "good job buddy" and put it on the fridge. I'm expecting these to get better as models improve, and perhaps the artistic progression will be there along with it...

Comment by jvanderbot 2 days ago

I call this "symbol drawing". Beginning artists do this. They think "This is a head, a head is round. This is where eyes go, eyes are shaped like this", and the whole thing ends up being a collage of symbols vs a representation of the space and experience of viewing a face. When I used to be OK at drawing, it was because I forced myself to use touch instead of sight to compose images. So weird to explain, but I'd feel the 3d to get the lighting and such better.

A lot of art that someone smarter than me told me to appreciate seems to follow the pattern of hitting the space and/or experience while minimizing the use of symbols. Impressionistic paintings esp avoid symbols IMHO, while bizzaro picassos abuse symbols outright and still hit the experience they are going for.

Comment by userbinator 3 days ago

Really odd to look at, considering if I saw any of these drawings from a human kid, I would say "good job buddy" and put it on the fridge.

The Grok ones in particular gave me that thought. Most of them really look like what a kid would do when given the same tools, while the other models' output have distinctly more "AI-ness" to them (for lack of a better term.)

Comment by MarcScott 2 days ago

Grok consistently draws images that, if I were its parent, I would refuse to put on the fridge door.

Comment by ralusek 3 days ago

What's interesting is that the way in which they're childish is actually extremely human. In fact, one of the ways that you're often taught to draw more realistically is to stop thinking of the concepts as icons you're drawing the outlines of, and instead sort of blur your eyes and see things as they are: hues and values. In other words, become a camera or a printer that has no idea what it's capturing or printing other than a grid of values. That is how you achieve realism.

The fact that it has clearly iconified these concepts in its mind and is tracing the outlines of the things it thinks/expects to go where is very human.

Comment by cj 3 days ago

Parrots can also sound extremely human, but it’s only mimicry.

How do we know the difference?

Comment by elcritch 3 days ago

Yes, but Parrots don't mimic the stages of learning speech development like children do. They just memorize a phrase.

These SOTA LLMs aren't trying to mimic existing children's drawings, but interestingly they're following somewhat similar progression that human children do as they develop.

Comment by diegof79 3 days ago

They don’t develop. We should stop anthropomorphizing LLMs.

AI labs are improving the ML techniques used to build better models; it’s a big difference.

Comment by perching_aix 2 days ago

That's very clearly what they're referring to also, or at least I certainly doubt this would be lost on them, or most anyone here, at this point.

Comment by diegof79 2 days ago

I imagine that it should be clear to everyone here now, until I read threads like this comparing child development to the advances in LLM building.

It’s like saying that a new M-Apple chip developed like a child. There is a clear tendency to give human like attributes to LLMs, probably caused by the terms used in ML: training, learning, etc.

Comment by perching_aix 2 days ago

I'm not so sure about this view.

For me, it was the same genre of comment as e.g. comments that call out how vision models tend to be structured surprisingly similarly to how the visual cortex is structured, even without specifically attempting to make that happen, and within the boundaries of our understanding of either things of course.

So not so much anthropomorphization, and more a recognition of the unintentional, and surprising, biomimicry.

Though to be honest, I'll be damned if any of the pictures in the post look anything like any child drawing I've ever seen, Grok's or any of the other models.

Comment by elcritch 2 days ago

No, it's more like you're being overly literal. As a field the technology is developing. Just like people might say "in the 60's integrated circuits developed quickly".

The parent wasn't referring to individual LLMs developing, but that overall new SOTA LLMs models are increasing in capabilities in ways similar to how human children do.

Comment by 3 days ago

Comment by ETH_start 3 days ago

Isn't LLM training, the development?

Comment by ipaddr 3 days ago

LLMs come complete they don't train on the fly based on interaction with the world. LLMs are wikipedia if they stopped allowing edits and all of the previous edits are training.

Comment by saidnooneever 3 days ago

there is infact RL for LLMs and its unlikely the frontier labs do not use it in some ways

Comment by tadfisher 3 days ago

The LLMs themselves are not doing the developing or learning.

Comment by ben_w 2 days ago

Both statements are true without contradiction.

A specific model develops as it passes through training.

AI labs change the architecture between models to allow them to surpass the previous models' best scores.

It's also going to keep being a hard sell to say "stop anthropomorphizing LLMs" when the models anthropomorphise themselves.

But besides that: Yeah, sure, they're not human, they're a cargo-cult mimicry of by and of minds that popped out of evolution doing gradient decent on intergenerational survivability. So what? Still interesting when the result of such cargo-culting incidentally echoes what our natural-selection-not-engineered moist electrochemistry happens to do.

Comment by richk449 3 days ago

If you want to know if something is a human or a parrot, you could give it a paint brush and ask it to paint the Mona Lisa.

Comment by saidnooneever 3 days ago

you can also use your eyes, hands or other senses :').

reminds me of those jokes?

what is difference between a parot and a human? silence rly? u cant tell the difference between a parot and a human? :')

Comment by CamperBob2 3 days ago

What's your parrot's take on the Jacobian Conjecture?

Comment by cj 2 days ago

My parrot doesn't have a take, but it sure can repeat one it heard on Reddit.

Comment by CamperBob2 2 days ago

So, you're saying that it was somebody on Reddit who disproved the conjecture, and not a "stochastic parrot?"

Comment by cj 2 days ago

You asked for my parrot's take, not how to disprove a conjecture.

Edit: And not a great example to use to prove human-ness considering humans didn't disprove the conjecture?

Comment by CamperBob2 2 days ago

Parrot analogies for frontier AI are so stupid that they prove human-ness by dint of being a mistake no thinking entity would make.

Comment by cj 2 days ago

Perhaps you can answer this for me.

Why do some people get very defensive when you suggest AI isn’t comparable to humans?

Like, where is it coming from?

Why do people seem to want to ascribe human traits on the tech? It’s just technology. Why do we need to draw parallels to biology? Is it actually helpful? My position is “no” because the parallels only seem similar, but fundamentally aren’t.

Comment by CamperBob2 2 days ago

Why do some people get very defensive when you suggest AI isn’t comparable to humans?

But you didn't do that, did you? You rattled off a goofy cliché about parrots that we've all heard a hundred times before. If that's all you have to contribute to the conversation, go contribute it somewhere else.

Comment by NichoPaolucci 2 days ago

Yes - precisely what I was getting at with the childish comment.

It felt a bit surreal to see a machines rendition of something that very closely maps to a younger human as they explore + understand more about the world. Scary, even (to me). This was the first time I’ve seen LLM output and thought “wow, maybe it is learning.”

Comment by danesparza 2 days ago

I took the same prompts to Gemini and was stunned by the results. The are completely different from the images shown in the article (and genuinely good art pieces that were generated).

Comment by jstanley 2 days ago

Were you using the Gemini image generator though? The article is about using LLMs to do drawing via tool calls.

Comment by 3 days ago

Comment by jnathsf 3 days ago

GPT 5.6 Sol had the best two drawings (rose and starry nights) but even more impressive was how efficient it was RE cost/time/tokens vs Fable (3.4M vs 14.6M / $7.74 vs $161!). OpenAI has quietly innovated around inference - this is will be a growing differentiator even against open models.

Comment by lubujackson 3 days ago

I do think they are going to stretch their lead in value if Anthropic doesn't wake up and stop YOLOing tokens. Kimi is an amazing achievement, but it has the same (or worse) kitchen sink approach as Fable.

At work, even if Fable is technically better I much prefer Sol because it is so much faster and concise.

Comment by theplumber 3 days ago

To me GPT is quite slow. I use it with max reasoning though but Fable at max as well

Comment by dajonker 3 days ago

For me personally, there is no point in using more reasoning than medium or sometimes high. It just gets so much slower and expensive, meaning that i have to wait longer to evaluate the results and provide feedback or continue to the next task.

Comment by ksd482 3 days ago

Grok! LOL!

Seriously, what's going on there ? Why is it so different from others? Is it just behind technologically/training wise or it's using something fundamentally different?

Comment by jorl17 3 days ago

Grok 4.5 is...something else.

It performs much better than composer2.5 (while being as fast). It's not Opus, but I think it's not that far off. Definitely better than sonnet for what I've been doing.

On the other hand, I think they probably heavily adapted the training data so that it really is extremely focused on code. I just recently ran my personal "poetry benchmark" on it (where I give it ~850 poems I've written over my life and ask it to comment the corpus as a whole), and it's whack. It tries to write in portuguese (most of the poems are portuguese) and code-switches constantly and mixes up words to the point of making what it writes almost unreadable (e.g. it writes stuff like "You can't QoS that that look for beast poems", in portuguese, all messed up). The quality of the analysis is also quite bad (I'd say it's definitely behind Sonnet).

So I really think they either threw away data that wasn't tied to coding so that they could fine-tune it to that, or somehow they've got such an unbalanced dataset that coding ends up dominating either way. To me, its disastrous performance in this drawing "competition" fits this narrative.

Comment by userbinator 3 days ago

"You can't QoS that" sounds like the title of a nerdcore rap song.

Comment by jorl17 3 days ago

Some of my poetry has clear IT jargon, but it's a very small portion of it (<1%). Some of my teenage poetry revolved a lot around the idea of wanting to become science, knowledge, and machine, and be rid of feeling altogether (to become an idea that has no body, or to be come the mathematical equations that define the world), with some very amateur odes written glorifying science and machine (a clear pastiche of Álvaro de Campos with a modern twist). But, again, this is not the majority of the work, far from it.

For some reason, OpenAI models, Gemini (and apparently Grok too), love to latch onto this and obsess over this idea that it's "programming poetry" or "poetry for the IT crowd". Often OpenAI and Gemini try to write the "equations of my poetry" (granted, I do write about a cyclical relationship between thinking, feeling and writing a lot, and I do have ONE poem which ends with a Q.E.D.).

I'm giving this context to say that it is very bizarre. It's as if they latch onto it and act as if it's a core or highly distinguished part of the poetry, when it really isn't. Anthropic models, on the other hand, absolutely do not do this, and have never done it.

I really don't understand why this happens. Maybe it's because it has a lot of portuguese, I don't know. And even though the "QoS" is clearly the wrong token being generated, I have had situations where gemini spoke of some phase of my poetry as the "Q&A part" (really, no joke...)

In any case, it's why it's my personal benchmark after all :D

Comment by saidnooneever 3 days ago

might also be U shaped memory issue did you try changing the order of your poems to see if it focusses on different ones? maybe it just happened to have those old ones in points it memory was focused at. (ofc its not good but would be an alternative idea to it focussing on IT things/code to produce such results)

Comment by jorl17 2 days ago

Yep, I did. It didn’t have any effect. Randomizing the poems has stopped having a significant effect in the last year or so on the analysis as a whole.

It still somewhat affects the way the AI looks at the poems, especially if it has to make lists of the “best”, but it is not a very pronounced effect nowadays. Maybe it gives some preference for earlier and latter poems but not a lot.

Back when I started doing it, these holes in the context were super obvious, just like you described. It would mixup entire periods and also hallucinate or forget about specific sections. That’s precisely why I started randomizing them as part of my experiments.

Nowadays the frontier models are much better and don’t really mix anything up.

Comment by giancarlostoro 3 days ago

They focused a little too much on Grok Imagine.

Comment by Ekshef 3 days ago

Did Elon tell some poor engineer to give grok a prompt injection for drawing,"make it look like one of my childhood drawings!" Just like the Tesla truck?

Comment by eth0up 3 days ago

The razor-wire at the bottom for Starry Night was clever, and very Grok. Really shows its military spirit.

Edit: I just don't see the point of redacting the Mona Lisa

Comment by yojo 3 days ago

I read the balls as “houses”, though the phallic spire emerging from them dead-center is also very on-brand

Comment by eth0up 3 days ago

I must admit, I have a natural tendency to overlook C&Bs, but solid catch -- it's there. Perhaps that explains the Mona Lisa redaction; Grok probably put more effort into that one.

Comment by vidarh 2 days ago

Redacting? Is that not just the model using the smudge tool it was given.

Comment by eth0up 2 days ago

Is this not someone being near infinitely obtuse?

HN, the only environment in the world where people will scientifically falsify a joke.

Comment by vidarh 2 days ago

If I saw something I thought resembled a joke I wouldn't have made the comment.

Comment by eth0up 2 days ago

ESL?

Allow me to explain the theory behind the humor...

Grok's art had heavy use of what strongly resembles a permanent marker with lateral-ish scribbles similar to what one might see when someone is trying to mark something as discarded. Documents, especially those procured by the US government when responding to FOIA requests, contain many 'strike-through' style black marks resembling bold permanent-marker lines, which obfuscate the words they wish the reader to remain ignorant of. In many cases, this is the entire document, especially where unredacted words might otherwise serve in any way to clarify the nature of the FOIA inquiry, i.e. answer the question.

Grok's art, resembled to me, and others potentially appreciating the joke, redactions more than art. In some forms of humor, the presenter of the joke feigns obtuseness, or a mildly self-deprecating, stupid perspective. But the stupidity of the perspective actually reveals a relevant concept, or idea.

Often, such jokes can be deciphered using common sense, or various razors, including both occam and hanlon, the former self-suggesting that the joker is probably not so stupid as to think Grok was really redacting anything, and the latter self-suggesting that there was probably no malice involved.

Comment by vidarh 2 days ago

It doesn't get any more funny when you try to explain something that wasn't funny in the first place.

Hence, I can't be bothered to read any of this.

Comment by eth0up 2 days ago

Then why pretend the question in the first place? Is that how you worked your way up to 40K karma tokens? Low hanging fruit harvests?

Look at the sequence: You inquired. I answered. Not finding something funny is not really a good reason to fake a question you never intended to really have answered. You are kind of admitting that you knew it was humor but decided to approach it deceptively, or manipulatively, as if I am the only one ever to make a joke that wasn't universally appreciated.

I think that makes you a prick, with a lot of karma-tokens to play with.

Comment by visiondude 3 days ago

the starry night one is soo funny

Comment by bdcravens 3 days ago

The Grok ones are amusing, almost comically bad. However, whenever I've tried to pass an image creation request to any of the Opus models, it's been far worse, like first week of using Microsoft Paint bad (while ChatGPT would create social media quality images using the same prompts)

Comment by vidarh 2 days ago

Opus doesn't have access to an image model, while ChatGPT does (it's not multi-modal, but it can prompt OpenAI's image model). So Opus by default is forced to write code to generate images, or generate SVG's as text. That's putting it in a worse situation than the linked article, where all the models were given rudimentary tools.

Give Opus access to an image model, and it will use that to create images just fine just like ChatGPT.

Comment by bibin765 2 days ago

A few months back, we had a use case to create some SVGs and added this skill: https://github.com/upbrew-tech/svg-creator-skill

I think using this skill, Claude drew it better than the examples given. https://claude.ai/share/bc0867dc-034f-4865-bb71-acb83bfa36a1

Comment by tylerrobinson 3 days ago

The Grok ones are so weird that they cross into uncanny and surreal. Truly bizarre and capable of eliciting feelings from me, if only bad feelings…

Comment by eth0up 3 days ago

Claude was clearly 'pushing back' on the coziness of the cabin. But I think it did best with the cat. Grok, I fear, is making a case for euthanasia. It's suffering and I think it would be cruel to let it continue. Someone pull the plug. .

Comment by hombre_fatal 3 days ago

GPT-5.6 Sol is the knock out here. Some of those results are really human/charming.

The rectangular smudge tool is a weird tool in the first place, but it's cute to see the models try to use it.

Comment by js8 3 days ago

I really like the cabin from Sol. It could be an art piece.

Comment by hombre_fatal 2 days ago

Its rose is also insane. Not just the shading of the flower but its clever use of the smudge tool to soften the background and then soften the vase so it looked more like glass.

And I like how its starry night decided to accentuate the church/steeple in the village backdrop.

Pretty cool how it did these with a scrappy little tool palette. What a fun idea.

Comment by js8 2 days ago

I found it funny that Fable reinterpreted the Starry Night as "an oil-well in a desert", while Grok went for a "Stalin statue in a prison camp" angle.

Comment by fastball 3 days ago

The most interesting result for me is that they apparently prompted the models to optimize for SSIM, but many of the models trend worse over time. I suppose because viewing the canvas always comes after drawing, and they didn't give "revert to previous" capability as part of the toolkit.

Which in turn kinda jives with my experience of using these models for code: to some extent they only seem to have a concept of "forward", which invariably leads to "write more code to fix previous problems created", rather than taking a step back and removing broken things entirely.

Comment by d1sxeyes 3 days ago

Stopping problem at work. I wonder if results would improve if we gave the agents an “undo” or “revert” tool.

Comment by ms7892 3 days ago

Fix this signup error please

https://www.tryai.dev/?error=server_error&error_code=unexpec...

Comment by NiloCK 3 days ago

For capabilities reference:

I made a lower effort but similar scaffold for LLMs to do iterative drawing in Nov 2024, with Sonnet 3.5 as the artist: https://paritybits.me/llm-drawing-with-eyes-open/

Quite a difference.

Comment by Athanase000 3 days ago

The way the LLM describes its own work at the end is hilarious.

Comment by traes 2 days ago

Perhaps there was something in the prompt or settings preventing this, but I'm surprised (and slightly disappointed) that none of the models approached this the way I would: download an image of the Mona Lisa, run all of the drawing functions many times to construct a forward model of the drawing implements, and attempt to solve some kind of explicit inverse problem through either ML or a classical algorithm to minimize some difference metric. Were they just restricted from running code or are the models unindustrious without a very specific prompt?

Comment by tulio_ribeiro 3 days ago

Weird choice of SSIM/RMSE. By feeding these back into the model, the agent is actively degrading its artistic output. This is only valid for the target reference I think?

Alternatively, I think much better results could be had by computing the cosine similarly using DINOv2 ONNX (@xenova/transformers).

My guess is that the results are greatly limited by relying on the current metrics.

Even better, attach toModelOutput directly to drawTool so it returns the rendered canvas image immediately upon drawing.

Comment by raybb 2 days ago

A year or so ago I was trying to get llms to make pixel art by asking them to create a 2d array of color codes. I'm no prompt guru but they all failed at anything beyond a basic shape. I tried at many resolutions from 16x16, 32x32, 64x64, and maybe a little larger but basically only then quit small ones were even recognizeable.

Comment by vunderba 2 days ago

There’s a related benchmark along these lines that uses an LLM with a discrete “instructive graphic language” to get models to generate Minecraft voxel structures with varying degrees of success.

https://github.com/Ammaar-Alam/minebench

Comment by dehrmann 2 days ago

This is picking the wrong tool for the job twice over. Half of it is seeing if text-to-image mechanics exist in general purpose LLMs, the other half is seeing if they can reason their way about generating the 2d array. Imagine asking this of yourself and doing it through just a train of thought. Yes, I could probably draw a basic smiley face in the color of your choice in a 16x16 without antialiasing in 15 minutes.

Comment by AlienRobot 3 days ago

The difference in cost is pretty incredible.

Comment by elcritch 3 days ago

Rather the difference of cost of Claude is pretty incredible. Then again, I guess they gotta make it rain while they can.

Comment by usef- 3 days ago

He mentions they used quite different methods to get to the end-result. I'd love to know how much changing the prompt impacts things: like telling claude to only draw and blend, similar to sol, etc.

There's likely not training data for this and Claude seemed to spend a lot of time blending the background.

Comment by MitziMoto 3 days ago

Grok's look like a truly disturbed child. Like the drawings from that kid in "The Ring".

Comment by LastTrain 3 days ago

Why include Grok when it is clearly not even in the same category as the other three?

Comment by sashank_1509 3 days ago

Where do these models even have this data to learn from. There must be massive computer use datasets? I have a hard time believing it’s emergent if it’s able to do something this good.

Comment by pavel_lishin 3 days ago

I think this is just an ad.

Comment by noodlescb 3 days ago

If it is, it's a horrible ad. My main takeaway is that all of them were god awful at image generation and Fable was 20x more expensive and still awful.

Comment by Legend2440 3 days ago

Well, your takeaway is bad. This isn't image generation, it's tool use with a digital paintbrush.

Comment by boomskats 3 days ago

And image tokenisation.

Comment by taf2 3 days ago

Would love to see the results if they had used /goal or similar

Comment by dizzard 3 days ago

I wonder how much better a harness could get for drawing

Comment by thomasfromcdnjs 3 days ago

I think a lot better

- the best image related stuff I've seen is where the harness is constantly cropping and looking closer at things (likely helps a lot for computer vision in general) - also it would be interesting because the harness could almost have its own "palette" as if it could play with blank squares and different strokes over lapping or blending before applying to the main canvas

Comment by aqfamnzc 2 days ago

I'm surprised that the harness wasn't better described here. Is the draw tool really just asking LLMs to provide a set of coordinates/parameters for generating shapes? Honestly, if so, the results are shockingly good. Think of how well a typical human artist would do with such a crude method of applying brushstrokes.

Comment by wolttam 3 days ago

Would be great to see Kimi here.

Comment by mbmbn 3 days ago

Shows about the same level of technical mastery that present days painters.

Comment by Razengan 3 days ago

Now this is a cool test. Sol's the best in each one, while Grok looks like the work of that person who "fixed" that Jesus painting..

Comment by fwip 3 days ago

Article is AI-written.

Comment by reactordev 3 days ago

Should have compared against FLUX Klein and Z-Image Turbo...

Comment by roguedemon 3 days ago

That was a good testcase. Pretty cool.

Comment by maxall4 3 days ago

I very much enjoyed Grok 4.5's rendition of the Mona Lisa as Elon Musk with tentacles.

Comment by SXX 3 days ago

Now add Deepseek, GLM and Kimi :-)

Comment by felixfurtak 3 days ago

I just tried with Deepseek and got this for the Mona Lisa

           .-""-.
          /      \
         |  _  _  |
         | (o)(o) |
         \   /\   /
          |  --  |
          |  \/  |
          |      |
         /   --   \
        /  /    \  \
       /  /      \  \
      (__/        \__)

It might not have been the most scientific test

Comment by fph 2 days ago

Looks more like Munch's The Scream to me.

Comment by ahoka 2 days ago

Any of those would be more relevant than Grok.

Comment by 3 days ago

Comment by flintapi 3 days ago

[flagged]

Comment by geroge_kyaw 3 days ago

Useless They are not image generation models.

Comment by Legend2440 3 days ago

That's what makes it an interesting challenge.

Comment by samuelknight 3 days ago

Not useless. LLMs are the most general purpose computer algorithms ever created. They are getting smarter and cheaper at a geometric rate. What is a bad idea today could have useful applications tomorrow.

Comment by munk-a 3 days ago

> cheaper at a geometric rate

Citation needed - my company is paying more than ever for code generation. I have no reason to believe (given anecdotes) that anyone finds themselves in the opposite situation.

Comment by beering 3 days ago

You’re paying more in total because the overall cost has gone down. This is normal economics when something useful gets cheaper. If you were to use original GPT-4 to do as much, you’d be paying an order of magnitude more.

Comment by ssl-3 3 days ago

It keeps getting cheaper and better -- for me.

I get a lot more use (read: cheaper per interaction) and much better quality results from the $20 that I spend on this stuff every month than I did several years ago.

(And several years before that, it was all essentially unobtanium.)

Comment by KolinFirz 3 days ago

Tell them to draw LeBron James or Heisenberg. Everyone will refuse.

Comment by leobg 3 days ago

I'm missing some context here.

First of all, this is a marketing piece, made for engagement.

Second, in production, you would adjust your approach to the model's strengths and weaknesses. For example, if you found that a model does not respond to the tools you are offering, then you would find a way to make it more responsive. Also, we know that different models handle prompts and system prompts differently. So comparing several models from different providers against exactly the same prompt is quite a naive way of exploring the capabilities of those models or comparing them to each other. You may have just happened to “speak the language“ of one model better than that of another. Why would that be the model’s fault and not yours?

Third, none of these results mean anything if you do not run each process on each model several times to check for consistency. Your model might just have gotten lucky. This point is even more true when you're talking about a progressive task where one step builds on the previous one. The model might just have painted itself into a corner.