Add new comment

How physics can help AI learn about the real world

It's always exciting to have a glimpse at new mathematics and technology as they take shape.  In this podcast we talk to Georg Maierhofer, from the University of Oxford, about an exciting new idea that is only just emerging  – physics informed neural networks  (PINNs for short) – where you add in the laws of physics to machine learning methods.  We have been able to sit in on a number of meetings of our colleagues from Maths4DL (the Mathematics for Deep Learning research group) as they explore this idea.   Georg explains why PINNs are a bit like learning golf, tells us the exciting opportunities and challenges, and why the key part to developing new ideas is getting the right people together at the right time.

You can find more about the machine learning and the some of the work that Maths4DL is doing here, including our recent podcast How does AI work?  and our collection Predicting the weather with artificial intelligence.

This content is part of our collaboration with the Mathematics for Deep Learning (Maths4DL) research programme, which brings together researchers from the universities of Bath and Cambridge, and University College London. Maths4DL aims to combine theory, modelling, data and computation to unlock the next generation of deep learning. You can see more content produced with Maths4DL here.

Maths4DL logo
You can listen to the podcast using the player above, and you can listen and subscribe to our podcast through Apple Podcasts, Spotify and through most other podcast providers via podbean.

Filtered HTML

  • Web page addresses and email addresses turn into links automatically.
  • Allowed HTML tags: <a href hreflang> <em> <strong> <cite> <code> <ul type> <ol start type> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.