For the first time ever, an artificial intelligence system can readily duplicate a Nobel Prize winning experiment, involving developing an ultra cold state of matter known as the Bose-Einstein condensate.

Named after physicists Satyendra Nath Bose and Albert Einstein, this state of matter was created when atomic particles reach temperatures almost close to absolute zero, or 0 Kelvin or equivalent to negative 459.6 degrees Fahrenheit. With this extremely low temperature, the atoms collect in the lowest possible energy state, that results in a giant matter wave.

Like Us on Facebook

This state of matter was predicted back in 1924 by Bose and Einstein however, this extreme state of matter was only achieved in a 1995 experiment that was also awarded the Nobel Prize for Physics in 2001. According to Paul Wigley of the Australian National University, it was surprising to study the machine that it can learn so much from scratch, and execute the same experiment in less than an hour. He also says that this AI algorithm played a crucial role since a simple computer program could have taken billions of years, almost the same age of the universe to comb through an infinite amount of combinations equations to figure this out.

In this new study, the researchers produced extremely cold gas and then trapped this inside a laser beam, where the AI system was put to the test. In order to achieve the exact temperature to just a few nanokelvins, which is one billion parts of one Kelvin, the most crucial part of this experiment that requires the utmost precision is creating the condensate.

This would entail that the AI would have to constantly monitor the different changes in parameters to allow the energetic state of atoms to escape, and keep the colder particles inside this laser beam.

This AI algorithm was proven to be a success to the team's surprise, as this is not programmed with any prior information about this process and was also able to from learn about and execute the experiment in under an hour using new methods that humans would not even thought of or resort to.

Wigley adds that the AI changed the laser's power up and down and measured it efficiently, coming up with complex approaches that humans would not have conjured up to achieve a colder state in experiments and also, execute more precise experiments as well.

This is the fastest recorded time that a Bose-Einstein condensate was created which means that a larger condensate can be developed in the future. The team also adds that the AI is also a cheaper way to carry out physics experiments than hiring a real physicist.

This new study is published in the journal, Nature Scientific Reports.

TagsAI, Artificial Intelligence, nobel prize, nobel prize physics, Bose-Einstein condensate, physics