Artificial intelligence applied to materials

This is a review of some artificial intelligence news that happened throughout the week.

Materials require very specific properties. The wide number of options to be tested is enormous and without the need to test them in the laboratory, simply simulating them takes a lot of time and resources. To that end, researchers from Carnegie Mellon University and the University of Calgary created AI algorithms that help them find the right material quickly and accurately. This is trained with known data making this a deep learning model.

Machine learning creating new drugs

Currently machine learning is used to analyze large amounts of data, helping biologists identify possible new drugs. This is useful in identifying how strongly a particular molecule is attached to a protein by the bacteria that cause tuberculosis.

Artist robots

A swarm of robots move on a canvas leaving the mark of their passage with respect to what the human artist indicates, creating art in real time. Artists and researchers come together to expand the frontiers of art. This was done by a studio that would help the artist to focus on the creation instead of worrying about the lines, these images are currently abstract and seem to have been created by a child. This can be the basis for swarm control in other fields of research.

Fewer neurons are better

There is currently a huge processing capacity that helps from autonomous vehicles to information searches. Research suggests that smaller neural networks can solve certain tasks more precisely. Artificial intelligence inspired by pinworm brains indicates a much better adaptation to noisy entrances and due to its simplicity its mode of operation can be explained in detail.

Speeding up the repair of software errors

Software bugs can be slow for programmers to resolve, taking weeks to find.
HangFix is ​​a software created by researchers at North Carolina State University, which detects and solves problems in seconds, this in order to detect bugs, diagnose problems and create the patch that corrects the error.

Thanks for reading, I hope you have informed yourself of something new. See you in the next edition.