By Gunnar Mathiason

Hybrid Office house rules

( A developing post.. stay tuned.. comment!) Yes, we have gotten quite good at Zoom meetings by now. As we soon get the chance of getting back to the physical office, there’s some pondered to do on how remote people can be included. Meetings! With digital meetings in the covid era, everyone joins on time….

By Gunnar Mathiason

AI Music and Creativity

I am attending “The 2020 Joint Conference on AI Music Creativity” this week. I was not aware of the advanced ML usage in music co-creation. There are many algorithmic tools available that actually do musical composition and real-time musical co-creation. I was stunned by the demo of the saxophonist that solo-dueled live together with an…

By Gunnar Mathiason

Learn how to write better

I have looked for good written sources of how to develop academic writing, targeting both PhD students and myself. I found the following three books and I do recommend them. Many people in the world who speak English professionally are non-native, and proper academic writing is hard but also needed. To me, the following reading…

By Gunnar Mathiason

Industrial Research School in Smart Industry

We’re applying for a new national Research School that is about Smart Industry. I believe that this is a good format for quickly boosting the collaboration and knowledge transfer between industry and academia, when it comes to making use of recent AI techniques. The KK foundation pays a part and employers of a company can…

By Gunnar Mathiason

My list of next research in AI

Yes, AI has come a long way, and many awesome things have been done recently. Many recent advancements have given very useful applications. Context-aware assistants, image and video interpretation, and Deep Learning for capturing very complex structures. Still, there are many things remaining on the to-do list before AI is here. Here’s my list of…

By Gunnar Mathiason

Probabilistic programming and the stan language

I’m learning the stan language for probabilistic programming, and what an experience! As an application case, we collaborate with Sandvik in the Swedish Metal project, to estimate the composition of recycled steel scrap. Probabilistic programming lends itself well for these kind of problems. What you do is that you define the problem as a statistical…