AI World Clocks • Hacker News Board

Re: AI World Clocks

#251

Post by leptons » Sat, Nov 15, 2025, 12:34 AM UTC

WJW wrote: post #219

bigfishrunning wrote: post #179

Earlier quoted context omitted.

How much engineering do prompt engineers do? Is it engineering when you add "photorealistic. correct number of fingers and teeth. High quality." to the end of a prompt? we should call them "prompt witch doctors" or maybe "prompt alchemists".

Well if it works consistently, I don't see any problem with that. If they have a clear theory of when to add "photorealistic" and when to add "correct number of wheels on the bus" to get the output they want, it's engineering. If they don't have a (falsifiable) theory, it's probably not engineering. Of course, the service they really provide is for businesses to feel they "do AI", and whether or not they do real engi…

>as relevant as if your favorite pornstars' boobs are real or not

This matters more than you might think.

Re: AI World Clocks

#252

Post by miohtama » Sat, Nov 15, 2025, 12:36 AM UTC

The new Turing time test

Re: AI World Clocks

#253

Post by nemomarx » Sat, Nov 15, 2025, 12:40 AM UTC

jdiff wrote: post #214

anigbrowl wrote: post #111

Earlier quoted context omitted.

I really like this. The broken ones are sometimes just failures, but sometimes provide intriguing new design ideas.

This same principle is why my favorite image generation model is the earlier models from 2019-2020 where they could only reliably generate soup. It's like Rorschach tests, it's not about what's there, it's about what you see in them. I don't want a bot to make art for me, sometimes I just want some shroom-induced inspirational smears.

I really miss that deepdream aesthetic with the dogs eyes popping up everywhere.

Re: AI World Clocks

#254

Post by earth2mars » Sat, Nov 15, 2025, 12:41 AM UTC

esafak wrote: post #181

earth2mars wrote: post #93

https://gemini.google.com/share/00967146a995 works perfectly fine with gemini 2.5 pro

how do you do that?

I used exactly the same prompt this site uses. Nothing else.

Re: AI World Clocks

#255

Post by bigbluedots » Sat, Nov 15, 2025, 12:45 AM UTC

I just realized I'm running late, it's almost -2!

More seriously, I'd love to see how the models perform the same task with a larger token allowance.

Re: AI World Clocks

#256

Post by Zopieux » Sat, Nov 15, 2025, 12:49 AM UTC

anon_cow1111 wrote: post #224

I'm having a hard time believing this site is honest, especially with how ridiculous the scaling and rotation of numbers is for most of them. I dumped his prompt into chatgpt to try it myself and it did create a very neat clock face with the numbers at the correct position+animated second hand, it just got the exact time wrong, being a few hours off. Edit: the time may actually have been perfect now that I account fo…

On the contrary, in my experience this is very typical of the average failure mode / output of early 2025 LLMs for HTML of SVG.

Re: AI World Clocks

#257

Post by bigbluedots » Sat, Nov 15, 2025, 12:50 AM UTC

Is there a "draw a pelican riding a bicycle" version?

Re: AI World Clocks

#258

Post by Workaccount2 » Sat, Nov 15, 2025, 12:50 AM UTC

chemotaxis wrote: post #148

Workaccount2 wrote: post #105

Earlier quoted context omitted.

The problem is more likely the tokenization of images than anything. These models do their absolute worst when pictures are involved, but are seemingly miraculous at generalizing with just text.

I wonder if it's because we mean different things by generalization. For text, "generalization" is still "generate text that conforms to all the usual rules of the language". For images of 13-hour clock faces, we're explicitly asking the LLM to violate the inferred rules of the universe. I think a good analogy would be asking an LLM to write in English, except the word "the" now means "purple". They will struggle to…

That's true, but I think humans would stumble a lot too (try reading old printed text from the 18fh cenfury where fhey used "f" insfead of t in prinf, if's a real frick fo gef frough).

However humans are pretty adept at discerning images, even ones outside the norm. I really think there is some kind of architectural block hampering transformers ability to really "see" images. For instance if you show any model a picture of a dog with 5 legs (a fifth leg photoshopped to it's belly) they all say there are only 4 legs. And will argue with you about it. Hell GPT-5 even wrote a leg detection script in python (impressive) which detected the 5 legs, and then it said the script was bugged, and modified the parameters until one of the legs wasn't detected, lol.

Re: AI World Clocks

#259

Post by anonzzzies » Sat, Nov 15, 2025, 12:55 AM UTC

Sonnet 4.5 does it flawless. Tried 8 times.

Re: AI World Clocks

#260

Post by hakcermani » Sat, Nov 15, 2025, 12:56 AM UTC

lanewinfield wrote: post #94

hi, I made this. thank you for posting. I love clocks and I love finding the edges of what any given technology is capable of. I've watched this for many hours and Kimi frequently gets the most accurate clock but also the least variation and is most boring. Qwen is often times the most insane and makes me laugh. Which one is "better?"

.. would you mind sharing the prompt .. in a gist perhaps .