[HN Gopher] The last six months in LLMs, illustrated by pelicans...
       ___________________________________________________________________
        
       The last six months in LLMs, illustrated by pelicans on bicycles
        
       Author : swyx
       Score  : 669 points
       Date   : 2025-06-08 07:38 UTC (15 hours ago)
        
 (HTM) web link (simonwillison.net)
 (TXT) w3m dump (simonwillison.net)
        
       | neepi wrote:
       | My only take home is they are all terrible and I should hire a
       | professional.
        
         | dist-epoch wrote:
         | Most of them are text-only models. Like asking a person born
         | blind to draw a pelican, based on what they heard it looks
         | like.
        
           | neepi wrote:
           | That seems to be a completely inappropriate use case?
           | 
           | I would not hire a blind artist or a deaf musician.
        
             | dist-epoch wrote:
             | The point is about exploring the capabilities of the model.
             | 
             | Like asking you to draw a 2D projection of 4D sphere
             | intersected with a 4D torus or something.
        
               | kevindamm wrote:
               | Yeah, I suppose it is similar.. I don't know their
               | diameters, rotations, nor the distance between their
               | centers, nor which two dimensions, so I would have to
               | guess a lot about what you meant.
        
             | namibj wrote:
             | It's a proxy for abstract designing, like writing software
             | or designing in a parametric CAD.
             | 
             | Most the non-math design work of applied engineering AFAIK
             | falls under the umbrella that's tested with the pelican
             | riding the bicycle. You have to make a mental model and
             | then turn it into applicable instructions.
             | 
             | Program code/SVG markup/parametric CAD instructions don't
             | really differ in that aspect.
        
               | neepi wrote:
               | I would not assume that this methodology applies to
               | applied engineering, as a former actual real tangible
               | meat space engineer. Things are a little nuanced and the
               | nuances come from a combination of communication and
               | experience, neither of which any LLM has any insight into
               | at all. It's not out there on the internet to train it
               | with and it's not even easy to put it into abstract terms
               | which can be used as training data. And engineering
               | itself in isolation doesn't exist - there is a whole
               | world around it.
               | 
               | Ergo no you can't just say throw a bicycle into an LLM
               | and a parametric model drops out into solidworks, then a
               | machine makes it. And everyone buys it. That is the hope
               | really isn't it? You end up with a useless shitty bike
               | with a shit pelican on it.
               | 
               | The biggest problem we have in the LLM space is the fact
               | that no one really knows any of the proposed use cases
               | enough and neither does anyone being told that it works
               | for the use cases.
        
               | dist-epoch wrote:
               | https://www.solidworks.com/lp/evolve-your-design-
               | workflows-a...
        
               | neepi wrote:
               | Yeah good luck with that. Seriously.
        
               | rjsw wrote:
               | I don't think any of that matters, CEOs will decide to
               | use it anyway.
        
               | neepi wrote:
               | This is sad but true.
        
             | __alexs wrote:
             | I guess the idea is that by asking the model to do
             | something that is inherently hard for it we might learn
             | something about the baseline smartness of each model which
             | could be considered a predictor for performance at other
             | tasks too.
        
             | dmd wrote:
             | Sorry, Beethoven, you just don't seem to be a match for our
             | org. Best of luck on your search!
             | 
             | You too, Monet. Scram.
        
             | simonw wrote:
             | Yeah, that's part of the point of this. Getting a state of
             | the art text generating LLM to generate SVG illustrations
             | is an inappropriate application of them.
             | 
             | It's a fun way to deflate the hype. Sure, your new LLM may
             | have cost XX million to train and beat all the others on
             | the benchmarks, but when you ask it to draw a pelican on a
             | bicycle it still outputs total junk.
        
               | dist-epoch wrote:
               | tried starting from an image:
               | 
               | https://chatgpt.com/share/684582a0-03cc-8006-b5b5-de51e5c
               | d89...
               | 
               | lol: https://gemini.google.com/share/4d1746a234a8
        
             | wongogue wrote:
             | Even Beethoven?
        
         | matkoniecz wrote:
         | it depends on quality you need and your budget
        
           | neepi wrote:
           | Ah yes the race to the bottom argument.
        
             | ben_w wrote:
             | When I was at university, they got some people from
             | industry to talk to us all about our CVs and how to do
             | interviews.
             | 
             | My CV had a stupid cliche, "committed to quality", which
             | they correctly picked up on -- "What do you mean?" one of
             | them asked me, directly.
             | 
             | I thought this meant I was focussed on being the best. He
             | didn't like this answer.
             | 
             | His example, blurred by 20 years of my imperfect human
             | memory, was to ask me which is better: a Porsche, or a go-
             | kart. Now, _obviously_ (or I wouldn 't be saying this),
             | Porsche was a trick answer. Less obviously is that _both_
             | were trick answers, because their point was that the
             | question was under-specified -- quality is the match
             | between the product and what the user _actually wants_ , so
             | if the user is a 10 year old who physically isn't big
             | enough to sit in a real car's driver's seat and just wants
             | to rush down a hill or along a track, none of "quality"
             | stuff that makes a Porsche a Porsche is of any relevance at
             | all, but what does matter is the stuff that makes a go-kart
             | into a go-kart... one of which is the affordability.
             | 
             | LLMs are go-karts of the mind. Sometimes that's all you
             | need.
        
               | neepi wrote:
               | I disagree. Quality depends on your market position and
               | what you are bringing to the market. Thus I would start
               | with market conditions and work back to quality. If you
               | can't reach _your_ standards in the market then you
               | shouldn 't enter it. And if your standards are poor, you
               | should be ashamed.
               | 
               | Go kart or porsche is irrelevant.
        
               | ben_w wrote:
               | > Quality depends on your market position and what you
               | are bringing to the market.
               | 
               | That's the point.
               | 
               | The market for go-karts does not support Porche.
               | 
               | If you bring a Porche sales team to a go-kart race,
               | nobody will be interested.
               | 
               | Porche doesn't care about this market. It goes both ways:
               | this market doesn't care about Porche, either.
        
         | keiferski wrote:
         | As the other guy said, these are text models. If you want to
         | make images use something like Midjourney.
         | 
         | Promoting a pelican riding a bicycle makes a decent image
         | there.
        
           | keiferski wrote:
           | * Prompting
        
         | GaggiX wrote:
         | An expert at _writing_ SVGs?
        
         | jug wrote:
         | Before that, you might ask ChatGPT to create a vector image of
         | a pelican riding a bicycle and then running the output through
         | a PNG to SVG converter...
         | 
         | Result:
         | https://www.dropbox.com/scl/fi/8b03yu5v58w0o5he1zayh/pelican...
         | 
         | These are tough benchmarks to trial reasoning by having it
         | _write_ an SVG file by hand and understanding how it's to be
         | written to achieve this. Even a professional would struggle
         | with that! It's _not_ a benchmark to give an AI the best tools
         | to actually do this.
        
           | YuccaGloriosa wrote:
           | I think you made an error there png is a bitmap format
        
             | sethaurus wrote:
             | You've misunderstood. The parent was making a specific
             | point -- if you want an SVG of a penguin, the easiest way
             | to AI-generate it is to get an image generator to create a
             | (vector-styled) bitmap, then auto-vectorize it to SVG. But
             | the point of this benchmark is that it's asking models to
             | create an SVG the _hard way_ , by writing its code
             | directly.
        
         | spaceman_2020 wrote:
         | My only take home is that a spanner can work as a hammer, but
         | you probably should just get a hammer
        
         | vunderba wrote:
         | This test isn't really about the quality of the image itself
         | (multimodals like gpt-image-1 or even standard diffusion models
         | would be far superior) - it's about following a spec that
         | describes _how to draw_.
         | 
         | A similar test would be if you asked for the pelican on a
         | bicycle through a series of LOGO instructions.
        
       | joshstrange wrote:
       | I really enjoy Simon's work in this space. I've read almost every
       | blog post they've posted on this and I love seeing them poke and
       | prod the models to see what pops out. The CLI tools are all very
       | easy to use and complement each other nicely all without trying
       | to do too much by themselves.
       | 
       | And at the end of the day, it's just so much fun to see someone
       | else having so much fun. He's like a kid in a candy store and
       | that excitement is contagious. After reading every one of his
       | blog posts, I'm inspired to go play with LLMs in some new and
       | interesting way.
       | 
       | Thank you Simon!
        
         | blackhaj7 wrote:
         | Same sentiment!
        
           | dotemacs wrote:
           | The same here.
           | 
           | Because of him, I installed a RSS reader so that I don't miss
           | any of his posts. And I know that he shares the same ones
           | across Twitter, Mastodon & Bsky...
        
       | nathan_phoenix wrote:
       | My biggest gripe is that he's comparing probabilistic models
       | (LLMs) by a single sample.
       | 
       | You wouldn't compare different random number generators by taking
       | one sample from each and then concluding that generator 5
       | generates the highest numbers...
       | 
       | Would be nicer to run the comparison with 10 images (or more) for
       | each LLM and then average.
        
         | puttycat wrote:
         | You are right, but the companies making these models invest a
         | lot of effort in marketing them as anything _but_
         | probabilistic, i.e. making people think that these models work
         | discretely like humans.
         | 
         | In that case we'd expect a human with perfect drawing skills
         | and perfect knowledge about bikes and birds to output such a
         | simple drawing correctly 100% of the time.
         | 
         | In any case, even if a model is probabilistic, if it had
         | correctly learned the relevant knowledge you'd expect the
         | output to be perfect because it would serve to lower the
         | model's loss. These outputs clearly indicate flawed knowledge.
        
           | ben_w wrote:
           | > In that case we'd expect a human with perfect drawing
           | skills and perfect knowledge about bikes and birds to output
           | such a simple drawing correctly 100% of the time.
           | 
           | Look upon these works, ye mighty, and despair:
           | https://www.gianlucagimini.it/portfolio-item/velocipedia/
        
             | jodrellblank wrote:
             | You claim those are drawn by people with "perfect knowledge
             | about bikes" and "perfect drawing skills"?
        
               | ben_w wrote:
               | More that "these models work ... like humans" (discretely
               | or otherwise) does not imply the quotation.
               | 
               | Most humans _do not_ have perfect drawing skills and
               | perfect knowledge about bikes and birds, they _do not_
               | output such a simple drawing correctly 100% of the time.
               | 
               | "Average human" is a much lower bar than most people want
               | to believe, mainly because most of us are average on most
               | skills, and also overestimate our own competence -- the
               | modal human has just a handful of things they're good at,
               | and one of those is the language they use, another is
               | their day job.
               | 
               | Most of us can't draw, and demonstrably can't remember
               | (or figure out from first principles) how a bike works.
               | But this also applies to "smart" subsets of the
               | population: physicists have https://xkcd.com/793/, and
               | there's this famous rocket scientist who weighed in on
               | rescuing kids from a flooded cave, they come up with some
               | nonsense about a submarine.
        
               | Retric wrote:
               | It's not that humans have perfect drawing skills, it's
               | that humans can judge their performance and get better
               | over time.
               | 
               | Ask 100 random people to draw a bike and in 10 minutes
               | and they'll on average suck while still beating the LLM's
               | here. Give em an incentive and 10 months and the average
               | person is going to be able to make at least one quite
               | decent drawing of a bike.
               | 
               | The cost and speed advantage of LLM's is real as long as
               | you're fine with extremely low quality. Ask a model for
               | 10,000 drawings so you can pick the best and you get a
               | marginal improvements based on random chance at a steep
               | price.
        
               | ben_w wrote:
               | > Ask 100 random people to draw a bike and in 10 minutes
               | and they'll on average suck while still beating the LLM's
               | here.
               | 
               | Y'see, this is a prime example of what I meant with
               | ""Average human" is a much lower bar than most people
               | want to believe, mainly because most of us are average on
               | most skills, and also overestimate our own competence".
               | 
               |  _An expert artist_ can spend 10 minutes and end up with
               | a brief sketch of a bike. You can witness this exact
               | duration yourself (with non-bike examples) because of a
               | challenge a few years back to draw the same picture in 10
               | minutes, 1 minute, and 10 seconds.
               | 
               | A _normal person_ spending as much time as they like gets
               | you the pictures that I linked to in the previous post,
               | because they don 't really know what a bike is. 45
               | examples of what normal people think a bike looks like:
               | https://www.gianlucagimini.it/portfolio-item/velocipedia/
               | 
               | > Give em an incentive and 10 months and the average
               | person is going to be able to make at least one quite
               | decent drawing of a bike.
               | 
               | Given mandatory art lessons in school are longer than 10
               | months, and yet those bike examples exist, I have no
               | reason to believe this.
               | 
               | > Ask a model for 10,000 drawings so you can pick the
               | best and you get a marginal improvements based on random
               | chance at a steep price.
               | 
               | If you do so as a human, rating and comparing images?
               | Then the cost is your own time.
               | 
               | If you automate it in literally the manner in this write-
               | up (pairwise comparison via API calls to another model to
               | get ELO ratings), ten thousand images is like $60-$90,
               | which is on the low end for a human commission.
        
               | zahlman wrote:
               | > A normal person spending as much time as they like gets
               | you the pictures that I linked to in the previous post,
               | because they don't really know what a bike is. 45
               | examples of what normal people think a bike looks like:
               | https://www.gianlucagimini.it/portfolio-item/velocipedia/
               | 
               | A normal person given the ability to consult a picture of
               | a bike while drawing will do much better. An LLM agent
               | can effectively refresh its memory (or attempt to look up
               | information on the Internet) any time it wants.
        
               | Retric wrote:
               | As an objective criteria what percentage include peddles
               | and a chain connecting one of the wheels? I quickly found
               | a dozen and stopped counting. Now do the same for those
               | LLM images and it's clear humans win.
               | 
               | > ""Average human" is a much lower bar than most people
               | want to believe
               | 
               | I have some basis for comparison. I've seen 6 years olds
               | draw better bikes than those LLM's.
               | 
               | Look through that list again the worst example does even
               | have wheels, multiple of them have wheels without being
               | connected to anything.
               | 
               | Now if you're arguing the average human is worse than the
               | average 6 year old I'm going to disagree here.
               | 
               | > Given mandatory art lessons in school are longer than
               | 10 months, and yet those bike examples exist, I have no
               | reason to believe this.
               | 
               | Art lessons don't cumulatively spend 10 months teaching
               | people how to draw a bike. I don't think I cumulatively
               | spent 6 months drawing anything. Painting, collage,
               | sculpture, coloring, etc art covers a lot and wasn't an
               | every day or even every year thing. My mandatory collage
               | class was art history, we didn't create any art.
               | 
               | You may have spent more time in class studying drawing,
               | but that's not some universal average.
               | 
               | > If you automate it in literally the manner in this
               | write-up (pairwise comparison via API calls to another
               | model to get ELO ratings), ten thousand images is like
               | $60-$90, which is on the low end for a human commission.
               | 
               | Not every one of those images had a price tag but one was
               | 88 cents, * 10,000 = 8,800$ just to make the image for a
               | test even at 4c/image your looking at 400$. Cheaper
               | models existed but fairly consistently had worse
               | performance.
        
               | simonw wrote:
               | The 88 cent one was the most expensive almost my an order
               | of magnitude. Most of these cost less than a cent to
               | generate - that's why I highlighted the price on the o1
               | pro output.
        
               | Retric wrote:
               | Yes, but if you're averaging cheap and expensive options
               | the expensive ones make a significant difference. Cheaper
               | is bound by 0 so it can't differ as much from the
               | average.
               | 
               | Also, when you're talking about how cheap something is,
               | including the price makes sense. I had no idea on many of
               | those models.
        
               | simonw wrote:
               | If you're interested, you can get cost estimates from my
               | pricing calculator site here: https://www.llm-
               | prices.com/#it=11&ot=1200
               | 
               | That link seeds it with 11 input tokens and 1200 output
               | tokens - 11 input tokens is what most models use for
               | "Generate an SVG of a pelican riding a bicycle" and 1200
               | is the number of output tokens used for some of the
               | larger outputs.
               | 
               | Click on different models to see estimated prices. They
               | range from 0.0168 cents for Amazon Nova Micro (that's
               | less than 2/100ths of a cent) up to 72 cents for o1-pro.
               | 
               | The most expensive model most people would consider is
               | Claude 4 Opus, at 9 cents.
               | 
               | GPT-4o is the upper end of the most common prices, at 1.2
               | cents.
        
               | Retric wrote:
               | Thanks
        
             | rightbyte wrote:
             | That blog post is a 10/10. Oh dear I miss the old internet.
        
           | cyanydeez wrote:
           | Humans absolutely do not work discretely.
        
             | loloquwowndueo wrote:
             | They probably meant deterministically as opposed to
             | probabilistically. Which also humans dont work like that :)
        
               | aspenmayer wrote:
               | I thought they meant discreetly.
        
           | bufferoverflow wrote:
           | > _work discretely like humans_
           | 
           | What kind of humans are you surrounded by?
           | 
           | Ask any human to write 3 sentences about a specific topic.
           | Then ask them the same exact question next day. They will not
           | write the same 3 sentences.
        
         | planb wrote:
         | And by a sample that has become increasingly known as a
         | benchmark. Newer training data will contain more articles like
         | this one, which naturally improves the capabilities of an LLM
         | to estimate what's considered a good ,,pelican on a bike".
        
           | cyanydeez wrote:
           | So what you really need to do is clone this blog post, find
           | and replace pelican with any other noun, run all the tests,
           | and publish that.
           | 
           | Call it wikipediaslop.org
        
             | YuccaGloriosa wrote:
             | If the any other noun becomes fish... I think I disagree.
        
           | criddell wrote:
           | And that's why he says he's going to have to find a new
           | benchmark.
        
           | viraptor wrote:
           | Would it though? There really aren't that many valid answers
           | to that question online. When this is talked about, we get
           | more broken samples than reasonable ones. I feel like any
           | talk about this actually sabotages future training a bit.
           | 
           | I actually don't think I've seen a single correct svg drawing
           | for that prompt.
        
         | simonw wrote:
         | It might not be 100% clear from the writing but this benchmark
         | is mainly intended as _a joke_ - I built a talk around it
         | because it 's a great way to make the last six months of model
         | releases a lot more entertaining.
         | 
         | I've been considering an expanded version of this where each
         | model outputs ten images, then a vision model helps pick the
         | "best" of those to represent that model in a further
         | competition with other models.
         | 
         | (Then I would also expand the judging panel to three vision
         | LLMs from different model families which vote on each round...
         | partly because it will be interesting to track cases where the
         | judges disagree.)
         | 
         | I'm not sure if it's worth me doing that though since the whole
         | "benchmark" is pretty silly. I'm on the fence.
        
           | ontouchstart wrote:
           | Very nice talk, acceptable by general public and by AI agent
           | as well.
           | 
           | Any concerns about open source "AI celebrity talks" like
           | yours can be used in contexts that would allow LLM models to
           | optimize their market share in ways that we can't imagine
           | yet?
           | 
           | Your talk might influence the funding of AI startups.
           | 
           | #butterflyEffect
        
             | threecheese wrote:
             | I welcome a VC funded pelican ... anything! Clippy 2.0
             | maybe?
             | 
             | Simon, hope you are comfortable in your new role of AI
             | Celebrity.
        
           | demosthanos wrote:
           | I'd say definitely do not do that. That would make the
           | benchmark _look_ more serious while still being problematic
           | for knowledge cutoff reasons. Your prompt has become popular
           | even outside your blog, so the odds of some SVG pelicans on
           | bicycles making it into the training data have been going up
           | and up.
           | 
           | Karpathy used it as an example in a recent interview:
           | https://www.msn.com/en-in/health/other/ai-expert-asks-
           | grok-3...
        
             | diggan wrote:
             | Yeah, this is the problem with benchmarks where the
             | questions/problems are public. They're valuable for some
             | months, until it bleeds into the training set. I'm certain
             | a lot of the "improvements" we're seeing are just
             | benchmarks leaking into the training set.
        
               | travisgriggs wrote:
               | That's ok, once bicycle "riding" pelicans become
               | normative, we can ask it for images of pelicans humping
               | bicycles.
               | 
               | The number of subject-verb-objects are near infinite. All
               | are imaginable, but most are not plausible. A
               | plausibility machine (LLM) will struggle with the
               | implausible, until it can abstract well.
        
               | diggan wrote:
               | > The number of subject-verb-objects are near infinite.
               | All are imaginable, but most are not plausible
               | 
               | Until there is enough unique/new subject-verb-objects
               | examples/benchmarks so the trained model actually
               | generalized it just like you did. (Public) Benchmarks
               | needs to constantly evolve, otherwise they stop being
               | useful.
        
               | demosthanos wrote:
               | To be fair, once it _does_ generalize the pattern then
               | the benchmark is actually measuring something useful for
               | deciding if the model will be able to product a subject-
               | verb-object SVG.
        
               | zahlman wrote:
               | I can't fathom this working, simply because building a
               | model that relates the word "ride" to "hump" seems like
               | something that would be orders of magnitude easier for an
               | LLM than visualizing the result of SVG rendering.
        
             | 6LLvveMx2koXfwn wrote:
             | I would definitely say he had no intention of doing that
             | and was doubling down on the original joke.
        
               | colecut wrote:
               | The road to hell is paved with the best intentions
               | 
               | clarification: I enjoyed the pelican on a bike and don't
               | think it's that bad =p
        
             | throwaway31131 wrote:
             | I'd say it doesn't really matter. There is no universally
             | good benchmark and really they should only be used to
             | answer very specific questions which may or may not be
             | relevant to you.
             | 
             | Also, as the old saying goes, the only thing worse than
             | using benchmarks is not using benchmarks.
        
           | fzzzy wrote:
           | Even if it is a joke, having a consistent methodology is
           | useful. I did it for about a year with my own private
           | benchmark of reasoning type questions that I always applied
           | to each new open model that came out. Run it once and you get
           | a random sample of performance. Got unlucky, or got lucky? So
           | what. That's the experimental protocol. Running things a
           | bunch of times and cherry picking the best ones adds human
           | bias, and complicates the steps.
        
             | simonw wrote:
             | It wasn't until I put these slides together that I realized
             | quite how well my joke benchmark correlates with actual
             | model performance - the "better" models genuinely do appear
             | to draw better pelicans and I don't really understand why!
        
               | pama wrote:
               | How did the pelicans of point releases of V3 and of R1
               | (R1-0528) do compared to the original versions of the
               | models?
        
               | MichaelZuo wrote:
               | I imagine the straightforward reason is that the "better"
               | models are in fact significantly smarter in some tangible
               | way, somehow.
        
               | more-nitor wrote:
               | I just don't get the fuss from the pro-LLM people who
               | don't want anyone to shame their LLMs...
               | 
               | people expect LLMs to say "correct" stuff on the first
               | attempt, not 10000 attempts.
               | 
               | Yet, these people are perfectly OK with cherry-picked
               | success stories on youtube + advertisements, while being
               | extremely vehement about this simple experiment...
               | 
               | ...well maybe these people rode the LLM hype-train too
               | early, and are desperate to defend LLMs lest their
               | investment go poof?
               | 
               | obligatory hype-graph classic: https://upload.wikimedia.o
               | rg/wikipedia/commons/thumb/9/94/Ga...
        
               | tuananh wrote:
               | until they start targeting this benchmark
        
               | simonw wrote:
               | Right, that was the closing joke for the talk.
        
               | jonstewart wrote:
               | It is funny to think that a hundred years in the future
               | there may be some vestigial area of the models' networks
               | that's still tuned to drawing pelicans on bicycles.
        
               | johnrob wrote:
               | Well, the most likely single random sample would be a
               | "representative" one :)
        
               | og_kalu wrote:
               | LLMs also have a 'g factor' https://www.sciencedirect.com
               | /science/article/pii/S016028962...
        
           | dilap wrote:
           | Joke or not, it still correlates much better with my own
           | subjective experiences of the models than LM Arena!
        
         | cyanydeez wrote:
         | I get my pelicans from google and my raw dogs from openAI,
         | while the best fundamental fascist ideologies are best sourced
         | from GrokAI.
        
         | qeternity wrote:
         | I think you mean non-deterministic, instead of probabilistic.
         | 
         | And there is no reason that these models need to be non-
         | deterministic.
        
           | skybrian wrote:
           | A deterministic algorithm can still be unpredictable in a
           | sense. In the extreme case, a procedural generator (like in
           | Minecraft) is deterministic given a seed, but you will still
           | have trouble predicting what you get if you change the seed,
           | because internally it uses a (pseudo-)random number
           | generator.
           | 
           | So there's still the question of how controllable the LLM
           | really is. If you change a prompt slightly, how unpredictable
           | is the change? That can't be tested with one prompt.
        
           | rvz wrote:
           | > I think you mean non-deterministic, instead of
           | probabilistic.
           | 
           | My thoughts too. It's more accurate to label LLMs as non-
           | deterministic instead of "probablistic".
        
         | mooreds wrote:
         | My biggest gripe is that he outsourced evaluation of the
         | pelicans to another LLM.
         | 
         | I get it was way easier to do and that doing it took pennies
         | and no time. But I would have loved it if he'd tried alternate
         | methods of judging and seen what the results were.
         | 
         | Other ways:
         | 
         | * wisdom of the crowds (have people vote on it)
         | 
         | * wisdom of the experts (send the pelican images to a few dozen
         | artists or ornithologists)
         | 
         | * wisdom of the LLMs (use more than one LLM)
         | 
         | Would have been neat to see what the human consensus was and if
         | it differed from the LLM consensus
         | 
         | Anyway, great talk!
        
           | zahlman wrote:
           | It would have been interesting to see if the LLM that Claude
           | judged worst would have attempted to justify itself....
        
         | timewizard wrote:
         | My biggest gripe is he didn't include a picture of an actual
         | pelican.
         | 
         | https://www.google.com/search?q=pelican&udm=2
         | 
         | The "closest pelican" is not even close.
        
       | anon373839 wrote:
       | Enjoyable write-up, but why is Qwen 3 conspicuously absent? It
       | was a really strong release, especially the fine-grained MoE
       | which is unlike anything that's come before (in terms of
       | capability and speed on consumer hardware).
        
         | Maxious wrote:
         | Cut for time - qwen3 was pelican tested too
         | https://simonwillison.net/2025/Apr/29/qwen-3/
        
         | simonw wrote:
         | Omitting Qwen 3 is my great regret about this talk. Honestly I
         | only realized I had missed it after I had delivered the talk!
         | 
         | It's one of my favorite local models right now, I'm not sure
         | how I missed it when I was reviewing my highlights of the last
         | six months.
        
       | qwertytyyuu wrote:
       | https://imgur.com/a/mzZ77xI here are a few i tried the models,
       | looks like the newer vesion of gemini is another improvement?
        
         | puttycat wrote:
         | The bicycle are still very far from actual ones.
        
           | simonw wrote:
           | I think the most recent Gemini Pro bicycle may be the best
           | yet - the red frame is genuinely the right shape.
        
             | layer8 wrote:
             | The pelican, on the other hand...
        
           | pjs_ wrote:
           | https://www.gianlucagimini.it/portfolio-item/velocipedia/
        
       | JimDabell wrote:
       | See also: The recent history of AI in 32 otters
       | 
       | https://www.oneusefulthing.org/p/the-recent-history-of-ai-in...
        
         | pbhjpbhj wrote:
         | That is otterly fantastic. The post there shows the breadth too
         | - both otters generated via text representations (in TikZ) and
         | by image generators. The video at the end, wow (and funny too).
         | 
         | Thanks for sharing.
        
       | bravesoul2 wrote:
       | Is there a good model (any architecture) for vector graphics out
       | of interest?
        
         | simonw wrote:
         | I was impressed by Recraft v3, which gave me an editable vector
         | illustration with different layers -
         | https://simonwillison.net/2024/Nov/15/recraft-v3/ - but as I
         | understand it that one is actually still a raster image
         | generator with a separate step to convert to vector at the end.
        
           | bravesoul2 wrote:
           | Now _that_ is a pelican on a bicycle! Thanks
        
       | dirtyhippiefree wrote:
       | Here's the spot where we see who's TL;DR...
       | 
       | > Claude 4 will rat you out to the feds!
       | 
       | >If you expose it to evidence of malfeasance in your company, and
       | you tell it it should act ethically, and you give it the ability
       | to send email, it'll rat you out.
        
         | ben_w wrote:
         | I'd say that's too short.
         | 
         | > But it's not just Claude. Theo Browne put together a new
         | benchmark called SnitchBench, inspired by the Claude 4 System
         | Card.
         | 
         | > It turns out nearly all of the models do the same thing.
        
           | dirtyhippiefree wrote:
           | I totally agree, but I needed you to post the other half
           | because of TL;DR...
        
         | yubblegum wrote:
         | I was looking at that and wondering about swatting via LLMs by
         | malicious users.
        
         | gscott wrote:
         | I am interested in this ratting you out thing. At some point
         | you have a video feed into AI from a Jarvis like headset
         | device, you walking down the street and cross the street in the
         | middle not at a sidewalk... does it rat you out? Does it make a
         | list of every crime no matter how small? Or just the big ones?
        
       | atxtechbro wrote:
       | Thank you, Simon! I really enjoyed your PyBay 2023 talk on
       | embeddings and this is great too! I like the personalized
       | benchmark. Hopefully the big LLM providers don't start gaming the
       | pelican index!
        
       | deadbabe wrote:
       | As a control, he should go on fiver and have a human generate a
       | pelican riding a bicycle, just to see what the eventual goal is.
        
         | gus_massa wrote:
         | Someone did this. Look at this sibling comment by ben_w
         | https://news.ycombinator.com/item?id=44216284 about an old
         | similar project.
        
           | zahlman wrote:
           | > back in 2009 I began pestering friends and random
           | strangers. I would walk up to them with a pen and a sheet of
           | paper asking that they immediately draw me a men's bicycle,
           | by heart.
           | 
           | Someone commissioned to draw a bicycle on Fiverr would not
           | have to rely on memory of what it should look like. It would
           | take barely any time to just look up a reference.
        
       | franze wrote:
       | Here Claude Opus Extended Thinking
       | https://claude.ai/public/artifacts/707c2459-05a1-4a32-b393-c...
        
         | ramesh31 wrote:
         | Single shot?
        
           | franze wrote:
           | 2 shot, first one did just generate the svg not the shareable
           | html page around it. in the second go it also worked on the
           | svg as i did not forbid it.
        
       | big_hacker wrote:
       | Honestly the metric which increased the most is the marketing and
       | astroturfing budget of the major players (OpenAI, Anthropic,
       | Google and Deepseek).
       | 
       | Say what you want about Facebook but at least they released their
       | flagship model fully open.
        
         | mdaniel wrote:
         | > model fully open.
         | 
         | uh-huh https://www.llama.com/llama4/license/
        
       | spaceman_2020 wrote:
       | I don't know what secret sauce Anthropic has, but in real world
       | use, Sonnet is somehow still the best model around. Better than
       | Opus and Gemini Pro
        
         | diggan wrote:
         | Statements like these are useless without sharing exactly all
         | the models you've tried. Sonnet beats O1 Pro Mode for example?
         | Not in my experience, but I haven't tried the latest Sonnet
         | versions, only the one before, so wouldn't claim O1 Pro Mode
         | beats everything out there.
         | 
         | Besides, it's so heavily context-dependent that you really need
         | your own private benchmarks to make head or tails out of this
         | whole thing.
        
       | wohoef wrote:
       | Quite a detailed image using claude sonnet 4:
       | https://ibb.co/39RbRm5W
        
       | landgenoot wrote:
       | If you would give a human the SVG documentation and ask to write
       | an SVG, I think the results would be quite similar.
        
         | diggan wrote:
         | Lets give it a try, if you're willing to be the experiment
         | subject :)
         | 
         | The prompt is "Generate an SVG of a pelican riding a bicycle"
         | and you're supposed to write it by hand, so no graphical
         | editor. The specification is here: https://www.w3.org/TR/SVG2/
         | 
         | I'm fairly certain I'd lose interest in getting it right before
         | I got something better than most of those.
        
           | mormegil wrote:
           | Did the testing prompt for LLMs include a clause forbidding
           | the use of any tools? If not, why are you adding it here?
        
             | diggan wrote:
             | The models that are being put under the "Pelican" testing
             | don't use a GUI to create SVGs (either via "tools" or
             | anything else), they're all _Text Generation_ models so
             | they exclusively use text for creating the graphics.
             | 
             | There are 31 posts listed under "pelican-riding-a-bicycle"
             | in case you wanna inspect the methodology even closer:
             | https://simonwillison.net/tags/pelican-riding-a-bicycle/
        
             | simonw wrote:
             | The way I run the pelican on a bicycle benchmark is to use
             | this exact prompt:                 Generate an SVG of a
             | pelican riding a bicycle
             | 
             | And execute it via the model's API with all default
             | settings, not via their user-facing interface.
             | 
             | Currently none of the model APIs enable tools unless you
             | ask them to, so this method excludes the use of additional
             | tools.
        
           | zahlman wrote:
           | > The colors use traditional bicycle brown (#8B4513) and a
           | classic blue for the pelican (#4169E1) with gold accents for
           | the beak (#FFD700).
           | 
           | The output pelican is indeed blue. I can't fathom where the
           | idea that this is "classic", or suitable for a pelican, could
           | have come from.
        
             | diggan wrote:
             | My guess would be that it doesn't see the web colors (CSS
             | color hexes) as proper hex triplets, but because of
             | tokenization it could be something dumb like
             | '#8B','451','3' instead. I think the same issue happens
             | around multiple special characters after each other too.
        
         | ramesh31 wrote:
         | >If you would give a human the SVG documentation and ask to
         | write an SVG, I think the results would be quite similar.
         | 
         | It certainly would, and it would cost at minimum an hour of the
         | human programmer's time at $50+/hr. Claude does it in seconds
         | for pennies.
        
       | adrian17 wrote:
       | > This was one of the most successful product launches of all
       | time. They signed up 100 million new user accounts in a week!
       | They had a single hour where they signed up a million new
       | accounts, as this thing kept on going viral again and again and
       | again.
       | 
       | Awkwardly, I never heard of it until now. I was aware that at
       | some point they added ability to generate images to the app, but
       | I never realized it was a major thing (plus I already had an
       | offline stable diffusion app on my phone, so it felt less of an
       | upgrade to me personally). With so much AI news each week, feels
       | like unless you're really invested in the space, it's almost
       | impossible to not accidentally miss or dismiss some big release.
        
         | azinman2 wrote:
         | Except this went very mainstream. Lots of turn myself into a
         | muppet, what is the human equivalent for my dog, etc. TikTok is
         | all over this.
         | 
         | It really is incredible.
        
           | thierrydamiba wrote:
           | The big trend was around the ghiblification of images. Those
           | images were everywhere for a period of time.
        
             | Jedd wrote:
             | Yeah, but so were the bored ape NFTs - none of these
             | ephemeral fads are any indication of quality, longevity,
             | legitimacy, or interest.
        
               | mrkurt wrote:
               | If we try really hard, I think we can make an exhaustive
               | list of what viral fads on the internet are not. You made
               | a small start.
               | 
               | none of these ephemeral fads are any indication of
               | quality, longevity, legitimacy, interest, substance,
               | endurance, prestige, relevance, credibility, allure,
               | staying-power, refinement, or depth.
        
               | Aurornis wrote:
               | 100 million people didn't sign up to make that one image
               | meme and then never use it again.
               | 
               | That many signups is impressive no matter what. The
               | attempts to downplay every aspect of LLM popularity are
               | getting really tiresome.
        
               | jodrellblank wrote:
               | I think it sounds far more likely that 100M people signed
               | up to poke at the latest viral novelty and create one
               | meme, than that 100M people suddenly discovered they had
               | a pressing long-term need for AI images all on the same
               | day.
               | 
               | Doesn't it?
        
               | gretch wrote:
               | It's neither of these options in this false dichotomy.
               | 
               | 100M people signed up and did at least 1 task. Then, most
               | likely some % of them discovered it was a useful thing
               | (if for nothing else than just to make more memes), and
               | converted into a MAU.
               | 
               | If I had to use my intuition, I would say it's 5% - 10%,
               | which represents a larger product launch than most
               | developers will ever participate in, in the context of a
               | single day.
               | 
               | Of course the ongoing stickiness of the MAU also depends
               | on the ability of this particular tool to stay on top
               | amongst increasing competition.
        
               | ben_w wrote:
               | While 100M signing up just for one pic is certainly
               | _possible_ , I note that several hundred million people
               | regularly share photographs of their lunch, so it is very
               | plausible that in signing up for the latest meme
               | generator they found they _liked_ the ability to generate
               | custom images of whatever they consider to be pretty
               | pictures every day.
        
               | baq wrote:
               | It's hard to think of a worse analogy TBH. My wife is
               | using ChatGPT to change photos (still is to this day),
               | she didn't use it or any other LLM until that feature
               | hit. It is a fad, but it's also a very useful tool.
               | 
               | Ape NFTs are... ape NFTs. Useless. Pointless. Negative
               | value for most people.
        
               | senthil_rajasek wrote:
               | "My wife is using ChatGPT to change photos (still is to
               | this day), she didn't use it or any other LLM until that
               | feature hit."
               | 
               | This is deja vu, except instead of ChatGPT to edit photos
               | it was instagram a decade ago.
        
               | jauntywundrkind wrote:
               | Applying some filters and adding some overlay text is
               | something some folks did, but there's such a massive
               | creative world that's opened up, where all we have to do
               | is ask.
        
               | baq wrote:
               | You either haven't tried it or are just trolling.
        
               | djhn wrote:
               | I tried it and I don't get it. What and where are the
               | legal usecases? What can you do with these low-resolution
               | images?
        
               | senthil_rajasek wrote:
               | I am contrasting how instagram filters gave users some
               | control and increased user base and how today editing
               | photos with LLMs is doing the same and pulling in a wider
               | user base.
        
               | micromacrofoot wrote:
               | they're not but I'm already seeing ai generated images on
               | billboards for local businesses, they're in production
               | workflows now and they aren't going anywhere
        
             | herval wrote:
             | They still are. Instagram is full of accounts posting gpt-
             | generated cartoons (and now veo3 videos). I've been
             | tracking the image generation space from day one, and it
             | never stuck like this before
        
               | simonw wrote:
               | Anecdotally, I've had several conversations with people
               | _way_ outside the hyper-online demographic who have been
               | really enjoying the new ChatGPT image generation - using
               | it for cartoon photos of their kids, to create custom
               | birthday cards etc.
               | 
               | I think it's broken out into mainstream adoption and is
               | going to stay there.
               | 
               | It reminds me a little of Napster. The Napster UI was
               | terrible, but it let people do something they had never
               | been able to do before: listen to any piece of music ever
               | released, on-demand. As a result people with almost no
               | interest in technology at all were learning how to use
               | it.
               | 
               | Most people have never had the ability to turn a photo of
               | their kids into a cute cartoon before, and it turns out
               | that's something they _really want_ to be able to do.
        
               | herval wrote:
               | Definitely. It's not just online either - half the
               | billboards I see now are AI. The posters at school. The
               | "we're hiring!" ad at the local McDonalds. It's 100x
               | cheaper and faster than any alternative (stock images,
               | hiring an editor or illustrator, etc), and most non
               | technical people can get exactly what they want in a
               | single shot, these days.
        
         | haiku2077 wrote:
         | Congratulations, you are almost fully unplugged from social
         | media. This product launch was a huge mainstream event; for a
         | few days GPT generated images completely dominated mainstream
         | social media.
        
           | derwiki wrote:
           | Not sure if this is sarcasm or sincere, but I will take it as
           | sincere haha. I came back to work from parental leave and
           | everyone had that same Studio Ghiblized image as their Slack
           | photo, and I had no idea why. It turns out you really can
           | unplug from social media and not miss anything of value: if
           | it's a big enough deal you will find out from another
           | channel.
        
             | dgfitz wrote:
             | I missed it until this thread. I think I'm proud of myself.
        
         | bufferoverflow wrote:
         | Have you missed how everyone was Ghiblifying everything?
        
           | adrian17 wrote:
           | I saw that, I just didn't connect it with newly added
           | multimodal image generation. I knew variations of style
           | transfer (or LoRA for SD) were possible for years, so I
           | assumed it exploded in popularity purely as a meme, not due
           | to OpenAI making it much more accessible.
           | 
           | Again, I was aware that they added image generation, just not
           | how much of a deal it turned out to be. Think of it like me
           | occasionally noticing merchandise and TV trailers for a new
           | movie without realizing it became the new worldwide box
           | office #1.
        
           | andrepd wrote:
           | Oh you mean the trend of the day on the social media
           | monoculture? I don't take that as an indicator of any
           | significance.
        
         | MattRix wrote:
         | To be clear: they already had image generation in ChatGPT, but
         | this was a MUCH better one than what they had previously. Even
         | for you with your stable diffusion app, it would be a
         | significant upgrade. Not just because of image quality, but
         | because it can actually generate coherent images and follow
         | instructions.
        
       | nowayno583 wrote:
       | That was a very fun recap, thanks for sharing. It's easy to
       | forget how much better these things have gotten. And this was in
       | just six months! Crazy!
        
       | mromanuk wrote:
       | The last animation is hilarious, represents very well the AI Hype
       | cycle vs reality.
        
       | bredren wrote:
       | Great writeup.
       | 
       | This measure of LLM capability could be extended by taking it
       | into the 3D domain.
       | 
       | That is, having the model write Python code for Blender, then
       | running blender in headless mode behind an API.
       | 
       | The talk hints at this but one shot prompting likely won't be a
       | broad enough measurement of capability by this time next year.
       | (Or perhaps now, even)
       | 
       | So the test could also include an agentic portion that includes
       | consultation of the latest blender documentation or even use of a
       | search engine for blog entries detailing syntax and technique.
       | 
       | For multimodal input processing, it could take into account a
       | particular photo of a pelican as the test subject.
       | 
       | For usability, the objects can be converted to iOS's native 3d
       | format that can be viewed in mobile safari.
       | 
       | I built this workflow, including a service for blender as an
       | initial test of what was possible in October of 2022. It took
       | post processing for common syntax errors back then but id imagine
       | the newer LLMs would make those mistakes less often now.
        
       | jfengel wrote:
       | It's not so great at bicycles, either. None of those are close to
       | rideable.
       | 
       | But bicycles are famously hard for artists as well. Cyclists can
       | identify all of the parts, but if you don't ride a lot it can be
       | surprisingly difficult to get all of the major bits of geometry
       | right.
        
         | mattlondon wrote:
         | Most recent Gemini 2.5 one looks pretty good. Certainly
         | rideable.
        
       | pier25 wrote:
       | Definitely getting better but even the best result is not very
       | impressive.
        
       | nine_k wrote:
       | Am I the only one who can't but see these attempts much like
       | attempts of a kid learning to draw?
        
         | Ygg2 wrote:
         | Yes. Kids don't draw that good of a line at the start.
         | 
         | Here is better example of start https://encrypted-
         | tbn0.gstatic.com/images?q=tbn:ANd9GcTfTfAA...
        
           | nine_k wrote:
           | Have you tried giving a kid a vector-drawing tool?
           | 
           | I did that to my daughter when she was not even 6 years old.
           | The results were somehow similar:
           | https://photos.app.goo.gl/XSLnTEUkmtW2n7cX8
           | 
           | (Now she's much better, but prefers raster tools, e.g.
           | https://www.deviantart.com/sofiac9/art/Ivy-with-riding-
           | gear-...)
        
       | djherbis wrote:
       | Kaggle recently ran a competition to do just this (draw SVGs from
       | prompts, using fairly small models under the hood).
       | 
       | The top results (click on the top Solutions) were pretty
       | impressive: https://www.kaggle.com/competitions/drawing-with-
       | llms/leader...
        
       | username223 wrote:
       | Interesting timeline, though the most relevant part was at the
       | end, where Simon mentions that Google is now aware of the
       | "pelican on bicycle" question, so it is no longer useful as a
       | benchmark. FWIW, many things outside of the training data will
       | pants these models. I just tried this query, which probably has
       | no examples online, and Gemini gave me the standard puzzle
       | answer, which is wrong:
       | 
       | "Say I have a wolf, a goat, and some cabbage, and I want to get
       | them across a river. The wolf will eat the goat if they're left
       | alone, which is bad. The goat will eat some cabbage, and will
       | starve otherwise. How do I get them all across the river in the
       | fewest trips?"
       | 
       | A child would pick up that you have plenty of cabbage, but can't
       | leave the goat without it, lest it starve. Also, there's no
       | mention of boat capacity, so you could just bring them all over
       | at once. Useful? Sometimes. Intelligent? No.
        
       | zahlman wrote:
       | > If you lost interest in local models--like I did eight months
       | ago--it's worth paying attention to them again. They've got good
       | now!
       | 
       | > As a power user of these tools, I want to stay in complete
       | control of what the inputs are. Features like ChatGPT memory are
       | taking that control away from me.
       | 
       | You reap what you sow....
       | 
       | > I already have a tool I built called shot-scraper, a CLI app
       | that lets me take screenshots of web pages and save them as
       | images. I had Claude build me a web page that accepts ?left= and
       | ?right= parameters pointing to image URLs and then embeds them
       | side-by-side on a page. Then I could take screenshots of those
       | two images side-by-side. I generated one of those for every
       | possible match-up of my 34 pelican pictures--560 matches in
       | total.
       | 
       | Surely it would have been easier to use a local tool like
       | ImageMagick? You could even have the AI write a Bash script for
       | you.
       | 
       | > ... but prompt injection is still a thing.
       | 
       | ...Why wouldn't it always be? There's no quoting or escaping
       | mechanism that's actually out-of-band.
       | 
       | > There's this thing I'm calling the lethal trifecta, which is
       | when you have an AI system that has access to private data, and
       | potential exposure to malicious instructions--so other people can
       | trick it into doing things... and there's a mechanism to
       | exfiltrate stuff.
       | 
       | People in 2025 actually need to be told this. Franklin missed the
       | mark - people today will trip over themselves to give up both
       | their security and their liberty for mere convenience.
        
         | simonw wrote:
         | I had the LLM write a bash script for me that used my
         | https://shot-scraper.datasette.io/ tool - on the basis that it
         | was a neat opportunity to demonstrate another of my own
         | projects.
         | 
         | And honestly, even with LLM assistance getting Image Magick to
         | output a 1200x600 image with two SVGs next to each other that
         | are correctly resized to fill their half of the image sounds
         | pretty tricky. Probably easier (for Claude) to achieve with
         | HTML and CSS.
        
           | zahlman wrote:
           | > And honestly, even with LLM assistance getting Image Magick
           | to output a 1200x600 image with two SVGs next to each other
           | that are correctly resized to fill their half of the image
           | sounds pretty tricky.
           | 
           | FWIW, the next project I want to look at after my current
           | two, is a command-line tool to make this sort of thing
           | easier. Likely featuring some sort of Lisp-like DSL to
           | describe what to do with the input images.
        
           | voiper1 wrote:
           | Isn't "left or right" _followed_ by rationale asking it to
           | rationalize it's 1 word answer - I thought we need to get AI
           | to do the chain of though _before_ giving it's answer for it
           | to be more accurate?
        
             | simonw wrote:
             | Yes it is - I would likely have gotten better results if
             | I'd asked for the rationale _first_.
        
       | Joker_vD wrote:
       | > most people find it difficult to remember the exact orientation
       | of the frame.
       | 
       | Isn't it D[?]L welded together? The bottom left and right
       | vertices are where the wheels are attached to, the middle bottom
       | point is where the big gear with the pedals is. The lambda is for
       | the front wheel because you wouldn't be able to turn it if it was
       | attached to a delta. Right?
       | 
       | I guess having my first bicycle be a cheap Soviet-era produced
       | one paid off: I spent _loads_ of time fidgeting with the chain
       | tension, and pulling the chain back onto the gears, so I guess I
       | had to stare at the frame way too much to forget even by today
       | the way it looks.
        
         | pbronez wrote:
         | There are a lot of structural details that people tend to gloss
         | over. This was illustrated by an Italian art project:
         | 
         | https://www.gianlucagimini.it/portfolio-item/velocipedia/
         | 
         | > back in 2009 I began pestering friends and random strangers.
         | I would walk up to them with a pen and a sheet of paper asking
         | that they immediately draw me a men's bicycle, by heart. Soon I
         | found out that when confronted with this odd request most
         | people have a very hard time remembering exactly how a bike is
         | made.
        
       | joshuajooste05 wrote:
       | Does anyone have any thoughts on privacy/safety regarding what he
       | said about GPT memory.
       | 
       | I had heard of prompt injection already. But, this seems
       | different, completely out of humans control. Like even when you
       | consider web search functionality, he is actually right, more and
       | more, users are losing control over context.
       | 
       | Is this dangerous atm? Do you think it will become more dangerous
       | in the future when we chuck even more data into context?
        
         | ActorNightly wrote:
         | Sort of. The thing is with agentic models, you are basically
         | entering probability space where it can do real actions in the
         | form of http requests if the statistical output leads it to it.
        
         | threeseed wrote:
         | I've had Cursor/Claude try to call rm -rf on my entire User
         | directory before.
         | 
         | The issue is that LLMs have no ability to organise their memory
         | by importance. Especially as the context size gets larger.
         | 
         | So when they are using tools they will become more dangerous
         | over time.
        
       | isx726552 wrote:
       | > I've been feeling pretty good about my benchmark! It should
       | stay useful for a long time... provided none of the big AI labs
       | catch on.
       | 
       | > And then I saw this in the Google I/O keynote a few weeks ago,
       | in a blink and you'll miss it moment! There's a pelican riding a
       | bicycle! They're on to me. I'm going to have to switch to
       | something else.
       | 
       | Yeah this touches on an issue that makes it very difficult to
       | have a discussion in public about AI capabilities. Any specific
       | test you talk about, no matter how small ... if the big companies
       | get wind of it, it _will_ be RLHF'd away, sometimes to the point
       | of absurdity. Just refer to the old "count the 'r's in
       | strawberry" canard for one example.
        
         | MattRix wrote:
         | This is why things like the ARC Prize are better ways of
         | approaching this: https://arcprize.org
        
         | simonw wrote:
         | Honestly, if my stupid pelican riding a bicycle benchmark
         | becomes influential enough that AI labs waste their time
         | optimizing for it and produce really beautiful pelican
         | illustrations I will consider that a huge personal win.
        
         | Choco31415 wrote:
         | Just tried that canard on GPT-4o and it failed:
         | 
         | "The word "strawberry" contains 2 letter r's."
        
       | zurichisstained wrote:
       | Wow, I love this benchmark - I've been doing something similar
       | (as a joke for and much less frequently), where I ask multiple
       | models to attempt to create a data structure like:
       | 
       | ``` const melody = [ { freq: 261.63, duration: 'quarter' }, // C4
       | { freq: 0, duration: 'triplet' }, // triplet rest { freq: 293.66,
       | duration: 'triplet' }, // D4 { freq: 0, duration: 'triplet' }, //
       | triplet rest { freq: 329.63, duration: 'half' }, // E4 ] ```
       | 
       | But with the intro to Smoke on the Water by Deep Purple. Then I
       | run it through the Web Audio API and see how it sounds.
       | 
       | It's never quite gotten it right, but it's gotten better, to the
       | point where I can ask it to make a website that can play it.
       | 
       | I think yours is a lot more thoughtful about testing novelty, but
       | its interesting to see them attempt to do things that they aren't
       | really built for (in theory!).
       | 
       | https://codepen.io/mvattuone/pen/qEdPaoW - ChatGPT 4 Turbo
       | 
       | https://codepen.io/mvattuone/pen/ogXGzdg - Claude Sonnet 3.7
       | 
       | https://codepen.io/mvattuone/pen/ZYGXpom - Gemini 2.5 Pro
       | 
       | Gemini is by far the best sounding one, but it's still off. I'd
       | be curious how the latest and greatest (paid) versions fare.
       | 
       | (And just for comparison, here's the first time I did it... you
       | can tell I did the front-end because there isn't much to it!)
       | https://nitter.space/mvattuone/status/1646610228748730368#m
        
         | dr_kretyn wrote:
         | I'm slightly confused by your example. What's the actual
         | prompt? Is your expectation that a text model is going to know
         | how to perform the exact song in audio?
        
           | zurichisstained wrote:
           | Ohhh absolutely not, that would be pretty wild - I just
           | wanted to see if it could understand musical notation enough
           | to come up with the correct melody.
           | 
           | I know there are far better ways to do gen AI with music,
           | this was just a joke prompt that worked far better than I
           | expected.
           | 
           | My naive guess is all of the guitar tabs and signal
           | processing info it's trained on gives it the ability to do
           | stuff like this (albeit not very well).
        
         | ojosilva wrote:
         | Drawbacks for using a pelican on a bicycle svg: it's a very
         | open-ended prompt, no specific criteria to judge, and lately
         | the svg all start to look similar, or at least like they
         | accomplished the same non-goals (there's a pelican, there's a
         | bicycle and I'm not sure its feet should be on the saddle or on
         | the pedals), so it's hard to agree on which is better. And,
         | certainly, having a LLM as a judge, the entire game becomes
         | double-hinged and who knows what to think.
         | 
         | Also, if it becomes popular, training sets may pick it up and
         | improve models unfairly and unrealistically. But that's true of
         | any known benchmark.
         | 
         | Side note: I'd really like to see the Language Benchmark Game
         | become a prompt based languages * models benchmark game. So we
         | could say model X excels at Python Fasta, etc. although then
         | the risk is that, again, it becomes training set and the whole
         | thing self-rigs itself.
        
       | NicoSchwandner wrote:
       | Nice post, thanks!
        
       | NohatCoder wrote:
       | If you calculate ELO based on a round-robin tournament with all
       | participants starting out on the same score, then the resulting
       | ratings should simply correspond to the win count. I guess the
       | algorithm in use take into account the order of the matches, but
       | taking order into account is only meaningful when competitors are
       | expected to develop significantly, otherwise it is just added
       | noise, so we never want to do so in competitions between bots.
       | 
       | I also can't help but notice that the competition is exactly one
       | match short, for some reason exactly one of the 561 possible
       | pairings has not been included.
        
         | simonw wrote:
         | Yeah, that's a good call out: Elo isn't actually necessary if
         | you can have every competitor battle every other competitor
         | exactly once.
         | 
         | The missing match is because one single round was declared a
         | draw by the model, and I didn't have time to run it again (the
         | Elo stuff was very much rushed at the last minute.)
        
       | m3047 wrote:
       | TIL: Snitchbench!
        
       | irthomasthomas wrote:
       | The best pelicans come from running a consortium of models. I use
       | pelicans as evals now.
       | https://x.com/xundecidability/status/1921009133077053462 Test it
       | using VibeLab (wip)
       | https://x.com/xundecidability/status/1926779393633857715
        
       ___________________________________________________________________
       (page generated 2025-06-08 23:00 UTC)