Artificial intelligence discovers amazing patterns in the destruction of the Earth’s biological mass

The new study used machine learning on fossil record to visualize the life story, showing the impact of major evolutionary events. This demonstrates the long-term evolutionary և ecological effects of the major events of destruction եկ speculation. Colors represent geological periods from 1 billion years ago, in yellow, to the current Quaternary, shown in green. The transition from red to blue marks the final Permian mass extermination, one of the most disruptive events in the fossil record. Loan J. Hoyal Cuthill և N. Guttenberg:

The idea that mass extinctions allow many new species to evolve is a central concept in evolution, but a new study that uses artificial intelligence to study fossil record is seldom true; there must be another explanation.

The famous work of Charles Darwin, About the origin of the species, ends with a beautiful summary of his theory of evolution. “In this view of life there is a greatness which has several powers, which were originally inspired by several horses or one. “Before this planet went cycling according to the fixed law of gravity, from such a simple beginning the most beautiful, the most wonderful, endless horses have evolved.”

In fact, scientists now know that most of the species that once existed became extinct. This extinction of the species has been roughly balanced by the emergence of new ones in the history of the Earth due to some major temporary imbalances that scientists call events of mass extinction. Scientists have long believed that mass extinctions create effective periods of species evolution or “radiation”, such as “creative extinction.” A new study by researchers at the Tokyo Institute of Technology, affiliated with the Earth Life Science Institute (ELSI), explored the coexistence of fossil species through machine learning.

Creative destruction is central to the classical notions of evolution. It seems clear that there are times when many species suddenly disappear, շատ many new species suddenly appear. However, radiation of a magnitude comparable to mass destruction, which this study, hereinafter referred to as mass radiation, has received far less analysis than destruction.

This study compared the effects of both extinction and radiation over the entire period for which there are fossils, the so-called Panerozoic Aeon. Phanerozoic (Greek for “obvious life”) represents the last 5 550 million վերջին ժամանակահատված 4.5 billion years of Earth պատմության’s significant period for archaeologists. Prior to that, most of the organisms that existed had bacteria that did not easily form fossils, so it is difficult to trace the original evolutionary record.

New research shows that creative extinction is not a good description of how species evolved or became extinct during the Phanerozoic period, suggests that there were far more remarkable periods of evolutionary radiation as life entered new evolutionary and ecological realms. what it was like in the Cambrian era. Explosion of animal diversity ային Carbon expansion of forest biomes. It is not known whether this was true of the previous միլի 3bn, where bacteria predominated, as the scarcity of information on such an old variety did not allow such an analysis to be made.

Paleontologists have uncovered some of the most brutal mass extinctions in the Phanerozoic fossil record. These mainly include the five major extinctions, such as the final Permian extinction, which estimated that more than 70% of the species became extinct. Biologists now believe that we can now enter the “Sixth Mass Extermination”, which, in their opinion, is mainly due to human activities, including hunting and the expansion of agriculture through land use. This is a common example of the mass extermination of the former “Big Five” Chalk– A secondary one (usually abbreviated as “KT” using the German chalk orthography), which seems to have originated when a meteorite struck Earth 65 million years ago, destroying non-avian dinosaurs.

Looking at the fossil record, scientists have come to the conclusion that mass extinction events produce particularly effective rays. For example, in the case of the extinction of KT dinosaurs, it is conventionally assumed that a desert has been created that allows mammalian organisms to colonize and “radiate”, allowing all new species of mammals to develop, eventually creating the basis for human evolution. In other words, if the “creative destruction” KT event had not taken place, we might not be here to discuss this issue.

The new study began with a random discussion in ELSI’s Agora, a common dormitory where ELSI scholars and visitors often dine and start new conversations. Two of the authors of the paper are evolutionary biologist Enn Jennifer Hoyal Kutil (now a researcher at the University of Essex in the UK) and physicist / mechanical learning expert Nicholas Gutenberg (now a Cross Labs researcher working with GoodAI of the Czech Republic, both ELSI). They were postgraduate doctoral students, and when the work began, they began to discuss the question of whether machine learning could be used to visualize or understand fossil record.

During the visit to ELSI, immediately before COVID-19: As the epidemic began to limit international travel, they feverishly sought to expand their analysis to investigate the link between disappearances and “radiation events.” These discussions allowed them to relate their new data to the breadth of existing ideas about mass destruction and radiation. They quickly found that evolutionary patterns identified through machine learning differed from key interpretations in traditional ways.

The team used a new application of machine learning to study the temporal coexistence of species in the Phanerozoic fossil record by examining more than one million entries in a massive, public database, including nearly two hundred thousand species.

The main author, Dr. Hoyal Kutil, said: “One of the most difficult aspects of understanding life story is the huge schedules and the involvement of the species. New applications of machine learning can help us visualize this information as read by humans. This means that we can, so to speak, keep half a billion years of evolution in our hands, gaining new insights from everything we see. ”

Using their objective methods, they found that cases of mass extermination of the “Big Five” previously discovered by archaeologists by machine learning methods were selected as the top 5% of significant disturbances in which the extinction was preceded by radiation or the opposite, as well as լրացուցիչ seven additional mass extinctions, two combined mass extinctions և fifteen mass emissions. In contrast to previous retellings that emphasized the power of posthumous radiation, this work found that more comparable mass radiation և annihilation was seldom combined over time, disproving the idea of ​​a causal link between them.

Co-author, Dr. Nicholas Gutenberg said: “The ecosystem is dynamic. You do not have to break an existing piece to get something new. ”

The team later discovered that radiation could cause real changes in existing ecosystems, an idea the authors call “destructive creation.” They found that in the Phanerozoic era, on average, almost all species that formed an ecosystem at any given time disappeared almost 19 million years later. But when mass extinction or radiation occurs, this rate of circulation is much higher.

This gives a new perspective on how the modern “Sixth Destruction” takes place. The Quaternary period, which began 2.5 million years ago, has witnessed repeated climate shocks, including abrupt changes in glaciers, a time when high-altitude positions on Earth were covered in ice. This means that the current “Sixth Extinction” is destroying already disturbed biodiversity, the authors estimate that it will take at least 8 million years for it to return to the long-term average of 19 million years. Dr. Hoyal Kutil says that “every destruction on our watch erases a species that could have existed for millions of years, making it difficult to replace the lost with the normal process of ’emerging new species.’ »

Reference. Spec enifer F. Hoyal Kutil, Nicholas Gutenberg և Graham E. Budd, “Effects of Spotting and Destruction Measured by the Evolutionary Degradation Clock”, 2020 December 9, Nature:,
DOI: 10.1038 / s41586-020-3003-4:

Related articles



Please enter your comment!
Please enter your name here

Share article

Latest articles

Double the number of known gravity lenses using artificial intelligence

Examples of gravity lenses in the DESI Heritage Study data. Credit: DESI Heritage Imaging Surveys / LBNL / DOE & KPNO / CTIO...

To rotate 2D materials

Illustration of the artistic concept of Spintronic computing. Scientists at Tsukuba University and the Institute of High-Pressure Physics are manufacturing a molybdenum disulfide transistor and...

Fine particles of veldfire smoke are more harmful than pollution by other sources

Researchers call for review of air quality monitoring guidelines to consider the sources of emissions. Researchers at the Scripps Institution of Oceanography at UC San...

Sudden turn in quantum physics may explain matter / antimatter imbalance

Lack of symmetry in cubes can not correct errors in quantum computing, but can explain matter / antimatter imbalance. The sudden turn in Kibble-Zurek theory...

The first people in New England may have shared the scene with woolly mammoths

A replica of a woolly mammoth (Mammuthus primigenius) at the BC Museum in BC, Victoria, British Columbia, Canada. The screen is from 1979...


Subscribe to stay updated.