Hacker News new | past | comments | ask | show | jobs | submit login
Over 300 New 'Nazca Lines' Geoglyphs Have Been Revealed by AI (thedebrief.org)
91 points by thunderbong 7 months ago | hide | past | favorite | 32 comments



"Orca with a knife". I have seen it before: https://photos.app.goo.gl/3sXkwYLboncsNYE8A


Where is that from?


A Nazca pot, here's another one and a picture of the whale line (article from 2019): https://www.antiguoperu.com/2019/11/cabezas-en-el-mar.html


That's insanely amazing.

Thanks for the share.


Some of them are more like people seeing the shape Jesus or a penis in clouds.


Yeah, I saw this yesterday and while most of these look legitimate, some of them are a real stretch.

It does make me wonder if these models suffer from pareidolia. If so, does that say anything about how similar they are to human intelligence? Or are they basically "inheriting" humanity's evolutionary biases as a function of their having been trained on human-generated content?


They can be confirmed by looking at the ground up close. The shapes are made removing pebbles from the desert floor and replacing them with differently-coloured pebbles.


The ground survey provides good confirmation that there was some human-created formation, though maybe what exactly they depict is harder to figure out.


Very interesting. These are the technical details I could infer from the paper

1. Collected data by flying aircraft over the area. Used a land classification mask to restrict the are to ~ 600 sq km

2. Make image patches of 11m by 11m. I believe there is some overlap in the patches. Sharpen the images for contrast.

3. The training data comes from previously known glyphs. Positive label patches are ones with a glyph. Negative label patches are randomly sampled from the vicinity of the glyph.

4. It looks like they fine tuned resnet 50 with these labels

5. Ran inference on other patches. They had false positives

6. Manually verified these AI predicted glyphs by ground surveys

I couldn't figure out how they drew the outlines in the pictures. I guess it was manually done


This is actually a really cool image recognition problem! I can imagine that an AI could actually be better at spotting these things than humans.


I have to say that I love these images. I love the earth images, the AI enhancements are cool, but above all I love the actual designs themselves. For me these designs are supreme examples of human creativity.


Don't let the desert racers find these new geoglyphs! "Desert Racers Demolish Art Carved by Ancient People in Chile": https://www.nytimes.com/2024/09/24/science/geoglyphs-atacama...


Can't wait for them to turn this tech loose on Mars photos. So many human faces waiting to be found.


Does this make more or less likely to be made by aliens in the minds of those that think so?


Naming that one figures the astronaut almost feel like someone is trolling those types. Looks nothing like an astronaut.


It is literally an amogus.. an among us crewmate in the exact format of the meme Like the ultimate case of pareidolia it has been spotted in trashcans, egyptian hieroglyphs as the god Medjed, now here. Wouldn’t be surprised if it was fake due to this meme alone but the sources I see all look completely legit, the image credit guy alone seems almost famous in wikipedia


I must say I don't follow you at all. The two eyes makes it look nothing like an astronaut or the amogus crewmate. The key visual there is a uniform visor, not two circles.


Primarily the two legs, no arms, and backpack Yeah you’re right there are two eyes which doesnt line up but the Internet went crazy for the amogus hieroglyph which also had eyes


It does look a bit like an owl


definitely looks like an owl to me also.


Who knows what a conspiracy theorist thinks? But honestly the fact that geoglyph construction techniques seem to have evolved over at least several hundred years to become ever clearer does seem to be a(n even stronger) point against. If it was aliens, wouldn't they have had the perfect method from the start? In fact, if it was actually aliens, why would they not just burn the images into the soil from orbit with a laser instead of laboriously piling up rocks?


>In fact, if it was actually aliens, why would they not just burn the images into the soil from orbit with a laser instead of laboriously piling up rocks?

They may have tried and accidentally destroyed all the other test planets with their death Star.


My recollection is that the main hypothesis among these crowds (where's the conspiracy?) is that the lines are human made, but in response to 'visiters'.


Could an ancient civilisation have their own Burning Man?


They did have ayahuasca, weeks of free time after the harvest, and a huge playa.


This reminds me of the ishahara color blindness test.


Would love to see the never-AI guys chime in about how'd they'd rather not have this than suffer the depredations of BigAI


I'm mostly unhappy with the current usage of LLMs. Not sure if I qualify as a "never-AI guy".

My main pet peeve is that AI is a blurry term almost to the point it has no definitions. It's basically a marketing term. AI has been applied, historically, to many trendy and novel facets of computer science, until they are not novel anymore and then they stop being AI. Expert systems, voice generation, image processing, genetic algorithms...

My take on the article: this is cool. I don't think I would call it AI though, if I could avoid it.


Correct, you already half name it. I'd call it image processing with pattern recognition


Any sufficiently advanced technology is indistinguishable from AI.


>”Humanoid”


looks fake, tbh - like "mary appeared on toast" fake




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: