Fri, May 17, 2013 - Page 9 News List

Developing artificial intelligence tools to build a curing machine

By H.T. Goranson and Beth Cardier

It does not stop there. When your neurons are replaced, they change. If you are newly in love (or experience a trauma) and a scent is associated with it, you may develop a heightened sensitivity to that scent. Your physio-cognitive apparatus evolves.

This process interests researchers enormously, because it is one of only two contexts in which neural regeneration occurs. If it were understood, it could lead to cures for many crippling diseases, both neurological (like Parkinson’s and Alzheimer’s) and degenerative (including those, like cancer, associated with aging).

One way to understand how this higher-level coherence emerges is to study it in a more accessible form. As it happens, there is a naturally occurring information structure in which the needed behaviors are easier to observe and explain. That structure can be found in stories. Stories are remarkable for the way that they allow us to comprehend new concepts and continue to make sense of them even when they involve numerous contexts and unexpected associations. We take the resulting narrative effects for granted: surprise endings, intriguing situations and a compulsion to read or watch until the finish.

If models of biology and artificial intelligence included these dynamics, they would demonstrate how unexpected elements can emerge from an evolving coherence.

Indeed, incorporating this principle into both fields could result in new diagnostic capabilities and new forms of individualized treatment, with a different therapy designed for each person.

An infection such as the coronavirus is currently countered with a vaccine tailored to block it. However, what if there was a way to “tune” bodies to reject all infections? For example, there are a few cases of natural immunity to HIV, but we are unable to understand why.

The challenge in biomedical and computational research is to model the dynamics between diverse processes at the level of whole systems. Once we can do this, we are much more likely to develop artificial intelligence tools that are capable of unexpected breakthroughs in understanding how the body supports and resists illness.

H.T. Goranson is the lead scientist of Sirius-Beta Corp and was a senior scientist with the US Defense Advanced Research Projects Agency. Beth Cardier is a researcher on narrative structure and artificial intelligence at the University of Melbourne, Australia.

Copyright: Project Syndicate

This story has been viewed 1702 times.
TOP top