Artificial Intelligence for photometric analysis in Archeology: New Discoveries in Nazca.


Recently I was reviewing news about the application of Artificial Intelligence to different fields, I could find news about its use in medicine, engineering, planning and management and even in possible uses in the evaluation of Human Resources within companies, but it called me the attention a news about its application in an old field of work that, as I recall, has always been good in using its auxiliary disciplines and supporting sciences to get new discoveries: The Field of Archaeological Research.

351665157-c466ae29f4cf14e2f6e80080753b0abc19db6e90c3034d52b0f8c2d46bcf0d27.jpeg

Source

 

I wrote about this topic in another blog that I publish on the Steem Blockchain, in the post: AI and Archeology: Making new discoveries in old research fields. Although I certainly wrote it with a somewhat more humorous introduction in that post.

Come on now with the central theme

Well, the news I saw was several, in different places (all in Spanish), but in general they were the same news:

The news was under the signature of Harold Moreno Luna.

In a way of summary, what was reported was that 143 new geoglyphs were discovered in the territory of the Pampas of Nasca thanks to a study by the University of Yamagata (Japan) applying IBM artificial intelligence (AI) technology. Akihisa Sakurai, an IBM engineer and member of the IBM Academy of Technology (IBM Research - Tokyo), stated:

“It is the same hardware technology behind the world's most powerful supercomputers, such as the Summit of the United States Department of Energy (which, since June 2018, is considered the fastest in the world, with a speed of 200 billion billions. of operations per second) or Sierra (of the National Nuclear Safety Administration, with a speed of 125 billion operations per second), ”he said.
Sakurai explained that they used Deep Learning or deep learning algorithms to detect figures on the ground. This technique uses several layers of information analysis, in which the ‘output’ information of one is the same as that of ‘input’ of another, allowing complex analysis and pattern recognition.
“Learning emulates the structure of the brain with neural networks implemented between‘ input ’and‘ output ’(information) of the Deep Learning model. By repeating this process several times, neural networks learn how they should connect to each other, based on learning parameters such as the lines that are part of the geoglyphs, ”he added. (Translated by my account)

The images captured by satellites, drones, aircraft and others were used to then confront any anomaly of the images and ensure that the system learned to differentiate the geographical accidents that had a human cause, that is, unnatural alterations of the anthropogenic type.

351665157-33c992b3b1ad4fcf1b503a6fe16e7924001b9074ec1b06a90005ecec7cb6b17b.jpeg

Source

The use of this technology demonstrated its usefulness by making dozens of new discoveries and giving information that was not available, thanks to the analysis and its subsequent verification in the field, the efficiency of the procedure could be evidenced and progress was made in the knowledge of this set of lines that make up the Nazca Lines, although we are still far from understanding the real meaning or meaning that this heritage of humanity has.

It's a good thing, but ...

Well, those who already know me, know of my tendency to be interested in technologies, but always keep in mind that one of the first fields of application of these are military and arms fields in general. So what could worry me about this news that is so academic?

The answer is simple, the AI ​​system learns to interpret objects or landforms from a means to differentiate those that are created by humans from those that are due to natural causes. I see it as obvious that if I place this AI in a satellite surveillance system, it could have a significant advantage in a war field where each side tries to hide its positions and the movements of its forces.

But also, if you load the AI ​​into a missile system and shoot them with the prerogative of modifying their targets if the information obtained on the ground flight allows them to identify a higher priority target than the one set at the time of launch.

All this and more are things that are already already under test or even already being implemented in the field, after all, I am used to the fact that every technology that reaches the civil world has been known and used by the military world for many years of advantage.


Post published also in the @project.hope community of Steem - https://beta.steemit.com/created/hive-175254

logo _20191218003205.png
Project #HOPE Website

How do you rate this article?

0


pedrobrito2004
pedrobrito2004

University professor in Venezuela. Living from the 20th to the 21st century


Curiosities & Varieties
Curiosities & Varieties

A blog to share post about Curiosities and Varieties of all kinds, while they are entertaining I will publish them :)

Send a $0.01 microtip in crypto to the author, and earn yourself as you read!

20% to author / 80% to me.
We pay the tips from our rewards pool.