The artificial neural network has identified a square structure in the triangular crater in the crater of the dwarf planet Ceres, which is appreciated. The results of this strange vision experiment conducted by neuropsychologists from the University of Cadiz made people question the application of artificial intelligence in the search for extraterrestrial life (SETI).
Although it is the largest object in the main asteroid belt, Ceres This is a dwarf planet. A few years ago, one of its craters became famous in one fell swoop: Profession, Where luminous points that caused various speculations were observed. When NASA’s Dawn probe got close enough and discovered that the origin of these bright spots was ice and salt released by the volcano, the mystery was solved.
Now from University of Cadiz Noticed one of the attractions, called Vinalia Faculae, And they are struck by areas that seem to be able to distinguish geometric shapes. This particularity led them to propose a strange experiment: comparing how humans and machines recognize planetary images.The ultimate goal is to analyze Artificial Intelligence (AI) Can help discover “Technology CompanyPossible extraterrestrial civilization.
“This is not just one of us, some people seem to recognize squares in Vinalia Faculae, so we see it as an opportunity to face humans and artificial intelligence in the cognitive task of visual perception, but not just any one. And it’s a challenging task associated with it Search for extraterrestrial life (SETI)Instead of just based on radio waves,” he explained Gabriel G. De la Torre.
From University of Cadiz Collected 163 volunteers who were not trained in astronomy to see what they enjoyed in the Occator image.Then they use Convolutional Neural Network, Its English abbreviation is), has received thousands of square and triangle images before training in order to be able to recognize them.
De la Torre emphasized: “Both humans and artificial intelligence detect square structures in images, but AI also recognizes triangles.” When showing the triangle option to humans, he claimed to also see the percentage of triangles. The square seems to Engraved on the triangle.
Another angle of the Vinalia Faculae area taken by the dawn probe (rotated 180º from above). The researchers studied the structure that appears in its central part, which is enlarged to the right, and in this structure is also marked (below) (indicated by numbers) the most commonly detected geometric shapes. /Original photo: NASA/JPL-Caltech/UCLA/MPS/DLR/IDA/PSI
These results are published in the journal Aerospace law, Which led the researchers to various conclusions: “On the one hand, despite being fashionable and having multiple applications, artificial intelligence may still confuse us and tell us that it has detected impossible or wrong things,” De La Torre said, so this Compromising its usefulness in the mission For example, looking for extraterrestrial technical structure. You must be careful about its implementation and use in SETI”.
“On the other hand, he added-if AI can recognize things that our brain cannot understand or accept, will it surpass our level of consciousness in the future and open the door to reality that we are not prepared for? ” What if the squares and triangles of Vinalia Faculae in Ceres are man-made structures? “
Finally, neuropsychologists pointed out that Artificial intelligence suffers from the same problems as creators: “Under human supervision, we should conduct more in-depth research on the influence of prejudice on its development.”
De La Torre concluded that, in fact, “we don’t know what it is, but the artificial intelligence detected in Vinalia Faculae is probably just a game. Lights and shadows“.
Gabriel G. De la Torre. “Does artificial intelligence dream of non-ground technology signatures?“. Aerospace law 167: 280-285, February 2020.
(In the same journal, the author published a previous study on the so-called cosmic gorilla effect in 2018).