Professor Still Invited to Present at the IBM Research Information Lens Workshop

Dr. Susanne Still has been invited to join a list of distinguished computer scientists and mathematicians at the First IBM Research Information Lens Workshop. The virtual workshop takes place from September 29 – October 2. The goal of the workshop is:

“Arguably, we are today in the midst of another information revolution, with the advent of neurons and qubits as new representation and processing elements for information. These advances, together with the exponential growth in memory and speed of conventional computing, have made it hazardous to conjecture any informational task at which humans will not be soon bested by computers. Viewing the world through an informational lens, and understanding constraints and tradeoffs such as energy and parallelism versus reliability and speed, will have profound consequences throughout technology and science. This includes not only mathematics and the natural sciences like physics and biology, but also social sciences such as psychology and linguistics. We aim to bring together leading researchers in science and technology from across the globe to discuss ideas and future research directions through the informational lens.”

The virtual workshop is open. Participants may register online. Please refer to the schedule for speaker names and times.

The abstract of Dr. Still’s talk appears below:

Title: Thermodynamic efficiency and predictive inference
Speaker: Suzanne Still (University of Hawaii at Mānoa)

Abstract: Learning from data about the surrounding world is not only essential for animal survival, but also a fundamental cornerstone of science. Observers are learning, model making, entities who use empirical inference. The search for a principled physical foundation of information processing and learning has a long history, spawning many developments that have become independent disciplines, including information theory and machine learning.

There are physical limits to information processing that neither living, nor artificial systems can circumvent. Tangible, physical advantages to certain strategies for representing and encoding data exist, and therefore, so does the possibility to derive machine learning methods from physics. I will discuss how this can be done using thermodynamic limits to information processing. Energy efficiency implies predictive inference, a strategy at the heart of machine learning.