Could AI eventually win a Nobel Prize?

AI is already transforming industries like banking, finance, film, and journalism, and researchers are now exploring how it could reshape their field—or even earn a Nobel Prize.

In 2021, Japanese scientist Hiroaki Kitano introduced the “Nobel Turing Challenge,” urging researchers to develop an “AI scientist” capable of conducting Nobel Prize-worthy research by 2050.

Some scientists are already striving to create such an AI, with this year’s Nobel laureates to be revealed between October 7 and 14.

The Emergence of Robot Scientists

Ross King, a professor of machine intelligence at Chalmers University in Sweden, noted that around 100 “robot scientists” already exist. In 2009, King and his colleagues unveiled “Robot Scientist Adam,” the first machine to independently make scientific discoveries.

“We developed a robot that autonomously discovered new science, generated original ideas, tested them, and confirmed their accuracy,” King told AFP.

This robot formed hypotheses on its own, designed experiments to test them, programmed lab robots to carry out those experiments, and then learned from the results to continue the process.

Unveiling New Gene Functions in Yeast Through Automation

“Adam” was programmed to study the biology of yeast and uncovered previously unknown gene functions in the organism. While the discoveries were considered “modest,” the robot’s creators emphasized they were still “not trivial.”

A second robot scientist, “Eve,” was later designed to investigate drug candidates for malaria and other tropical diseases.

According to King, robot scientists already possess certain advantages over human scientists. “They’re less expensive, work around the clock, and are meticulous in recording every detail,” he explained.

However, King acknowledged that AI is still far from becoming a Nobel-caliber scientist, stating it would need to be “much more intelligent” and capable of grasping “the bigger picture.”

Inga Strumke, an associate professor at the Norwegian University of Science and Technology, reassured that the scientific profession remains secure for now. “Machines are nowhere near taking over science,” she told AFP.

However, Strumke acknowledged that “it’s not impossible” and that AI is already influencing how science is conducted and will continue to do so.

Revolutionizing Protein Structure Prediction Through AI

One example is AlphaFold, an AI model developed by Google DeepMind that predicts the 3D structure of proteins based on their amino acid sequences. “We knew there was a connection between amino acids and the protein’s final 3D shape, and machine learning helped us find it,” Strumke explained, adding that the complexity of these calculations was beyond human capabilities. “We now have a machine doing something humans couldn’t.”

However, AlphaFold also highlights a limitation of current AI models, like neural networks. While they excel at processing vast amounts of data and producing answers, they struggle to explain why those answers are correct. Although AlphaFold’s predictions of over 200 million protein structures are “incredibly useful,” Strumke noted, they don’t “teach us anything about microbiology.”

AlphaFold’s Potential Impact on Nobel Recognition

According to Strumke, science is about understanding the universe, not just “making the correct guess.” However, the pioneering work done by AlphaFold has led to speculation that its creators might soon be in line for a Nobel Prize.

In 2023, Google DeepMind’s director John Jumper and CEO Demis Hassabis were awarded the prestigious Lasker Award, and analytics group Clarivate lists them as top contenders for the 2024 Nobel Prize in Chemistry, to be announced on October 9.

David Pendlebury, head of Clarivate, noted that while a 2021 paper by Jumper and Hassabis has been widely cited, it’s unusual for the Nobel jury to recognize such recent work—most honored discoveries are decades old. Still, he believes it won’t be long before AI-assisted research earns a Nobel Prize. “I’m confident that within the next decade, Nobel Prizes will be awarded to work aided by AI,” Pendlebury told AFP.

Read the original article on: Phys Org

Read more: Navigating the Complexities of Generative AI in Education