Artificial intelligence is already transforming sectors ranging from banking and finance to film and journalism, and scientists are exploring how AI could revolutionize their fields—or even earn a Nobel Prize. In 2021, Japanese scientist Hiroaki Kitano introduced the 'Nobel Turing Challenge,' inviting researchers to develop an 'AI scientist' capable of conducting autonomous research worthy of a Nobel Prize by 2050. Some scientists are already working diligently to create an AI colleague that could potentially win a Nobel, with this year's laureates set to be announced between October 7 and 14.
There are approximately 100 'robot scientists' in existence, according to Ross King, a professor of machine intelligence at Chalmers University in Sweden. In 2009, King and his colleagues published a paper introducing 'Robot Scientist Adam'—the first machine to independently make scientific discoveries. 'We built a robot that discovered new science on its own, generated novel scientific ideas, tested them, and confirmed their correctness,' King explained to AFP. The robot was designed to form hypotheses autonomously, design experiments to test these hypotheses, and even program laboratory robots to conduct the experiments, learning from the process and repeating it. 'Adam' was tasked with exploring the inner workings of yeast and discovered 'functions of genes' previously unknown in the organism. The creators of the robot scientist noted that while the discoveries were 'modest,' they were 'not trivial' either.
A second robot scientist, named 'Eve,' was later developed to study drug candidates for malaria and other tropical diseases. According to King, robot scientists already offer several advantages over human scientists. 'It costs less money to do the science, they work 24/7,' he said, adding that they are also more meticulous in recording every detail of the process. However, King acknowledged that AI is far from being a Nobel-worthy scientist. For that, they would need to be 'much more intelligent' and capable of 'understanding the bigger picture.'
Inga Strumke, an associate professor at the Norwegian University of Science and Technology, believes that the scientific profession is safe for now. 'The scientific tradition is nowhere near being taken over by machines anytime soon,' she said. However, she added that 'it doesn't mean that it's impossible,' noting that AI is definitely having and will continue to have an impact on how science is conducted. One example is AlphaFold, an AI model developed by Google DeepMind, which predicts the three-dimensional structure of proteins based on their amino acid sequence. 'We knew there was some relation between the amino acids and the final three-dimensional shape of the proteins... and then we could use machine learning to find it,' Strumke explained. She noted that the complexity of such calculations was too daunting for humans. 'We kind of have a machine that did something no humans could do,' she said.
However, Strumke also pointed out a weakness of current AI models like neural networks. They excel at processing vast amounts of information and providing answers, but struggle to explain why those answers are correct. While the over 200 million protein structures predicted by AlphaFold are 'extremely useful,' they 'don't teach us anything about microbiology,' Strumke said. For her, science seeks to understand the universe and is not merely about 'making the correct guess.' Despite this, the groundbreaking work of AlphaFold has led some to consider its creators as potential Nobel Prize winners. Google DeepMind's director John Jumper and CEO and co-founder Demis Hassabis were awarded the prestigious Lasker Award in 2023. Analytics group Clarivate, which tracks potential Nobel science laureates, lists the pair among the top picks for the 2024 Prize in Chemistry, to be announced on October 9.
David Pendlebury, head of the research group, acknowledges that while a 2021 paper by Jumper and Hassabis has been cited thousands of times, it would be unusual for the Nobel jury to award work so soon after publication—most discoveries honored date back decades. However, he is confident that it won't be long before research aided by AI wins the most coveted science prize. 'I'm sure that within the next decade there will be Nobel Prizes that are somehow assisted by computation and computation these days is more and more AI,' Pendlebury said.
