In an amazing scientific development, a new artificial intelligence tool named AI co-scientist developed by Google was able to solve a problem that took scientists a decade to solve in just two days, and this achievement highlights the huge potential of artificial intelligence in accelerating the pace of scientific research and providing innovative solutions to complex challenges.
Artificial intelligence overcomes 10 years of scientific research:
Jose benades and his colleagues at Imperial College London have spent a whole decade trying to understand the mechanism by which some types of superbugs acquire resistance to antibiotics, an increasingly serious global health challenge that claims millions of lives annually.
But the amazing surprise happened when the team presented the new AI co-scientist tool from Google-an artificial intelligence tool designed specifically for collaboration with researchers – the same question that took them a decade to solve, and in just two days, the tool provided the same answer that the team came up with in their results, which have not yet been published.
Benadis was surprised by this quick and accurate result, which prompted him to communicate with Google to verify that they did not have access to his search, the company denied this, stressing that the tool relied on publicly available information. The researchers published their results on February 19 in the bioRxiv database, which means that they have not yet undergone peer review.
Thiago Dias Da Costa, a lecturer in bacterial pathology at Imperial College London and co-author of the study, said: “Our results show that artificial intelligence has the ability to collect all the available evidence and guide us towards the most important questions and experimental designs, and if this tool proves to be as effective as we expect, it will make a quantum leap in our field, helping to avoid dead ends and accelerate the pace of discoveries by unprecedented.
Using artificial intelligence to combat superbugs:
An amazing achievement.. Artificial intelligence solves a decade-long scientific mystery in just two days
Antimicrobial resistance (AMR) is a dangerous phenomenon that is constantly worsening, as infectious microbes, such as bacteria, viruses, fungi and parasites, acquire resistance to antibiotics, which makes essential medicines ineffective, leading to an exacerbation of diseases and an increase in mortality rates.
Antimicrobial resistance is known as the silent pandemic, and it is one of the biggest health threats facing humanity, as the overuse and misuse of antibiotics in medicine and agriculture accelerate their spread and development.
According to a report released by the Centers for Disease Control and Prevention (CDC) in 2019, drug-resistant bacteria caused at least 1.27 million deaths globally that year, and of those, there were about 35,000 deaths in the United States alone.
This means that deaths in the United States from this problem increased by 52% compared to the antimicrobial resistance report issued by the Centers for Disease Control in 2013, highlighting the speed at which this threat is developing.
In their quest to understand the mechanisms of resistance of superbacteria, the benades team focused on a specific type of this bacterium, a family of viruses that infect bacteria and are known as phage-inducible chromosomal islands that form the protein envelope of the virus particle (cf-PICIs). These viruses have aroused the interest of researchers for their unique ability to infect a variety of types of bacteria, which suggests an unknown penetration mechanism.
Researchers have developed a hypothesis that these viruses acquire this ability by borrowing tails from other viruses that attack bacteria, and use these tails to inject the genetic material of the virus into the host bacterial cell, and experiments have proved this hypothesis correct, revealing an innovative breakthrough mechanism in the horizontal gene transfer process that the scientific community was not aware of before, and this mechanism allows viruses to switch tails between some, which increases their ability to infect different types of bacteria.
Before publishing their results, the team decided to test the validity of their discovery using Google'S new AI co-scientist tool, and just two days later, the artificial intelligence made multiple suggestions, one of which was identical to the answer the team came up with, and this result confirmed the ability of artificial intelligence to analyze complex data and provide valuable insights in the field of scientific research.
Commenting on these results, benadis, professor of microbiology at Imperial College London, said: “the algorithm has actually been able to review the available evidence, analyze the probabilities, ask questions, design experiments and propose the same hypothesis that it took us years of rigorous scientific research to arrive at, but in a fraction of the time.
This means that artificial intelligence has not limited itself to analyzing data, but has gone beyond this to simulating the process of scientific thinking, which opens up new horizons for accelerating scientific discoveries.
Challenges of using artificial intelligence in scientific research:
The researchers pointed out that relying on artificial intelligence from the very beginning would not have dispensed with conducting practical experiments, but it would have significantly accelerated the process of reaching the hypothesis, saving them years of work. This highlights the complementary role of artificial intelligence in scientific research, helping to guide efforts and set priorities, but it does not replace experimentation and scientific verification.
With these and other promising results, the use of artificial intelligence in the scientific field is still being debated, as many studies have proven that research supported by artificial intelligence is unrepeatable-that is, other researchers cannot replicate practical results or reproduce them using the same method or procedures involves outright fraud.
To reduce these problems and maximize the benefits that artificial intelligence can bring to research, scientists propose to develop tools to detect the misuse of artificial intelligence and create ethical frameworks for evaluating the accuracy of results, and this requires cooperation between researchers, developers, and regulators to ensure the responsible and ethical use of artificial intelligence in scientific research.