Recommended Free Tools
Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Artificial intelligence has not decoded the Nazca Lines like an ancient cipher. But a collaboration between Yamagata University and IBM Research used AI to help identify and prioritize archaeological targets, leading field researchers to confirm 303 previously unknown figurative geoglyphs in Peru in just six months.
The discovery nearly doubled the known number of figurative Nazca geoglyphs and provided stronger evidence that different types of designs may have served different social and ritual purposes. It was a major archaeological breakthrough—but “solved” is an overstatement.
What are the Nazca Lines?
The Nazca geoglyphs are enormous designs created in Peru’s coastal desert by removing dark surface stones and exposing the lighter ground underneath. The wider Nazca region is a UNESCO World Heritage site.
The term “Nazca Lines” is often used broadly. It can refer to long straight lines, geometric shapes, trapezoids, and recognizable figures. The 2024 study focused on figurative geoglyphs—designs representing recognizable forms such as humans, animals, and domesticated camelids—not every line or geometric feature in the landscape.
#1 Best Overall
The desert’s dry conditions helped preserve these designs for centuries. Their existence is well documented, but their precise purposes remain debated. Archaeologists have proposed connections to ritual activity, movement through the landscape, social gatherings, and relationships between communities and the environment. Popular explanations, including extraterrestrial theories, are not supported by the archaeological evidence presented in this research.
How AI helped archaeologists find them
The project, published in Proceedings of the National Academy of Sciences on September 23, 2024, used deep-learning image analysis to examine aerial imagery and identify locations that might contain overlooked geoglyphs. The research was conducted by Yamagata University’s Institute of Nazca and IBM Research. The open-access study is also listed in PubMed.
Its workflow was closer to intelligent archaeological triage than autonomous discovery:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Researchers assembled and processed high-resolution aerial imagery.
- A deep-learning object-detection system analyzed known visual patterns associated with geoglyphs.
- The system generated and ranked possible locations.
- Archaeologists selected promising candidates for field investigation.
- Teams visited the locations, checked the terrain, and confirmed genuine geoglyphs.
- The researchers recorded each feature’s shape, location, context, and relationship to nearby paths and other designs.
That human verification is essential. The AI did not excavate sites, independently authenticate them, date them, or determine what they meant. It helped researchers search a huge landscape more efficiently.
Rank #2
What was discovered?
During six months of field survey, researchers confirmed 303 new relief-type figurative geoglyphs. Their addition nearly doubled the known inventory of figurative geoglyphs in the Nazca region.
“Nearly doubled” does not mean that AI doubled every Nazca Line, every geometric feature, or the total area covered by the Nazca geoglyph system. It refers specifically to the number of known figurative geoglyphs.
The newly documented designs included human-related motifs and domesticated camelids, among other figurative forms. The central importance of the discovery is not a particular list of animals. It is the much larger, better-mapped dataset researchers can now use to study how the geoglyph landscape was organized.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errorsWhy had the figures been missed?
Many relief-type geoglyphs are relatively small, faint, weathered, or difficult to distinguish from the surrounding terrain. The Nazca landscape is also extensive. Manually examining every aerial image and then deciding where to send field teams is slow and expensive.
AI’s advantage was scale. It could systematically screen image archives and highlight candidates that deserved specialist attention. In supporting material, Yamagata University described an earlier deep-learning workflow as approximately 21 times faster than manual image analysis. The later study also reported a 16-fold increase in the rate of discovery in its particular comparison.
Those figures describe specific workflows and comparison methods. They should not be treated as a universal claim that AI makes all archaeological research 16 or 21 times faster.
What the new map suggests about the Nazca landscape
The expanded evidence helped researchers distinguish between two broad categories of geoglyphs:
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →| Type | Common characteristics | Researchers’ interpretation |
|---|---|---|
| Line-type geoglyphs | Generally larger; often associated with long lines, trapezoids, and wild-animal imagery | Likely connected to community-level ritual activity |
| Relief-type geoglyphs | Generally smaller; more often depict humans or domesticated camelids and occur near winding trails | May have been viewed by individuals or small groups moving through the landscape |
This distinction matters because it challenges the idea that every Nazca design had one identical purpose. The researchers argue that the placement, size, imagery, and relationship to paths point toward different viewing practices and social functions.
However, these are archaeological interpretations, not direct records of what every Nazca artist or community intended. Distribution patterns can provide strong evidence, but they do not reveal the complete meaning of an individual figure.
What AI did not solve
The research did not conclusively answer:
- Why the Nazca people created the entire geoglyph system.
- Whether all figures served the same religious, social, or practical purpose.
- How rituals involving the geoglyphs were organized.
- What each individual motif represented to its makers.
- How the designs changed across different periods.
- Who created particular geoglyphs or how much labor each required.
It is therefore inaccurate to say that AI “understood the meaning” of the drawings. The technology improved detection. Archaeologists then used the enlarged body of evidence to develop and test more persuasive explanations.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.The limits of AI in archaeology
Automated image analysis is powerful, but it is not neutral or infallible. Natural terrain, shadows, erosion, vehicle tracks, and modern disturbances can produce false positives. A model trained on known examples may also favor designs that resemble previously documented forms and overlook genuinely unusual ones.
Image resolution, lighting, topography, erosion, and uneven image coverage can affect what the system detects. Human choices still shape the outcome: researchers select imagery, define training examples, set thresholds, evaluate candidates, and decide which locations receive fieldwork.
Even a field-confirmed geoglyph may require additional dating, contextual analysis, and comparison with nearby features. Archaeological data must also be handled in accordance with Peru’s heritage laws and the protection needs of vulnerable sites. Better mapping can support conservation by identifying areas threatened by erosion, vehicles, construction, or other damage, but discovery alone does not guarantee protection.
The broader lesson
The Nazca project illustrates where AI is most useful in archaeology: searching, ranking, organizing, and comparing evidence at a scale that would overwhelm a small human team.
It does not replace remote sensing specialists, GIS analysis, field archaeologists, historians, or local institutions. Instead, it connects large image collections with limited fieldwork resources. The result is more efficient evidence gathering, followed by human examination and interpretation.
So the accurate headline is not that AI solved one of archaeology’s biggest puzzles. It is that AI helped archaeologists find enough additional evidence to reshape the puzzle.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
