Scientists spent 10 years in study and was going to solve a problem with superbacteria in two days

0

An Artificial Intelligence (AI) tool has managed to solve, in just two days, a problem that microbiologists took ten years to unravel.

Professor José R. Penadés, from Imperial College London, and his team spent years studying why some superbacteria are resistant to antibiotics. As a test, he asked the “co-scientist” tool created by Google about the problem. To his surprise, AI reached the same conclusion as the survey in just 48 hours.

Penades was impressed as his study had not yet been published. This means that AI did not have access to discoveries from public sources, making the result even more surprising.

See also:

The researchers sought to understand how some superbacteria arise. The hypothesis was that they can develop “tails” made of viruses, which facilitates their propagation between different species.

Penades compared the phenomenon to a key that allows these bacteria to “change” from one host to another. Until then, this theory was exclusive to the team and had not been disclosed.

To test AI, the teacher submitted the issue to the Google tool. In just two days, the system suggested hypotheses – and the first of them confirmed exactly the theory developed by scientists over the past decade.

In addition to validating the initial hypothesis, AI suggested four other possibilities. All made sense, and one of them was something the researchers had not yet considered. Now the team is exploring this new line of investigation.

Penades stressed that the tool not only reproduced the study accurately, but brought new perspectives. “They provided right hypotheses, and one of them was something we had never thought of. Now we are working on it, ”he said.

The advance shows how artificial intelligence can revolutionize science by saving years of research and opening unexpected ways for new discoveries.