Laws of Quantum Mechanics Learned by AI Algorithm

The rapid development in the field of artificial intelligence (AI) is pretty impressive. Globally, different industries are investing in AI due to the benefits it brings and could bring to their various endeavors. Technology sites like have been closely following the progress in AI technology.

Recently, news broke out that AI algorithms can now be taught to learn the laws of quantum mechanics. A group of researchers at the University of Warwick, the University of Luxemburg, and the Technical University of Berlin have developed a groundbreaking AI method that can predict molecular wave functions and digital properties of molecules.

Chemists can use this deep machine learning method to accelerate the simulation efforts within the design of drug molecules or new compounds. If you want to learn more about this technological and scientific breakthrough and its importance, read on below.

A Revolutionary Study

In the mind-boggling field of quantum chemistry, chemists use artificial intelligence and machine learning algorithms to calculate vital chemical properties and foretell experimental outcomes. But, for the sake of accuracy, AIs must have a clear understanding of the fundamentals of quantum mechanics.

According to an interdisciplinary crew of chemists, computer scientists, and physicists from the universities mentioned above, machine learning frameworks in the past were not quite capable of making accurate quantum predictions.

It’s not to say that previous AI algorithms were not useful, but the thing is that they failed to pinpoint some of the most vital characteristics of quantum chemistry. Specifically, previous AI frameworks have neglected to consider in these trials how many changing factors are needed to mark out a particular state of a system.

Quantum mechanics commonly allows states to exist and not exist at the same time. If scientists utilize degrees of freedom, they can know and understand how to describe a system with a certain degree of accuracy and utility.

Failing to account for such degrees of freedom, former AI algorithms have described these quantum chemistry trials in more classical tensor, vector, and scalar fields. This knowledge gap means that more time and effort are required for calculation and prediction.

Thanks to this new study, an AI method can now look deeper into the quantum structure and behavior of molecules with much accuracy. The research team has designed a model that will describe these molecular structures and behaviors in a faster and more quantumly accurate form of ground-state wavefunctions.

Reinhard Maurer, an assistant professor at the University of Warwick’s Department of Chemistry and one of the authors of the study, said that their AI algorithm is quantum-savvy and flexible, which makes it an essential tool for quantum chemistry.

According to Maurer, the study has been conducted for three years, and it required expert knowledge in computer science to develop an AI algorithm that can determine the structure and behavior of wave functions accurately. Physics and chemistry know-how are also vital to represent and process quantum chemical data that can be managed by the algorithm.

A Synergy of Quantum Chemistry and Machine Learning

The research team said in a statement that their deep machine learning model known as SchNOrb enables them to estimate molecular orbits with quantum chemical accuracy. As such, the framework also helps them predict the electronic structure of molecules and chemically interpret the reaction dynamics of said molecules.

This deep machine learning algorithm shows promising capabilities that will help chemists design effective “purpose-built molecules” that can be used in the field of medicine and various industries.

However, every one of us should understand that the standard rules that we make in the visible world don’t always apply in quantum mechanics. After all, quantum mechanics work at the sub-atomic level.

Although SchNOrb can be a useful framework in quantum chemistry, the high number of atomic orbitals it can process also makes it open to a high number of prediction errors. And if many prediction errors are accumulated over time, chances are there will be congestion in the prediction process.

The authors of the study said that they would need to learn more about the algorithm and improve it to prevent the problem mentioned above.

Still, the researchers are confident that the breakthrough they just made will open the way for a more effective synergy between quantum-savvy AIs and quantum chemists.


Artificial intelligence has done the world an excellent service even though there are a lot of detractors that often roll out a long list of AIs’ supposed absurd threats to humanity. Right now, a new AI algorithm is being hailed as an essential tool in advancing the field of quantum mechanics.

This AI model known as SchNOrb can dig deeper into the structure and behavior of molecules, which, in turn, speeds up the simulation efforts in the design of drug molecules and new compounds. Of course, this innovative discovery is proof that artificial intelligence can bring about a positive impact if it’s incorporated into various scientific fields.

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button