FutureMed: Artificial Intelligence to Fuel Nanotech and Other Breakthroughs

Posted in Medical Computing by Brian Buntz on February 5, 2013

“The computers are in control. We just live in their world.” That quote, attributed to supercomputer pioneer Danny Hillis in a Wired magazine article was meant to be a reflection of the world we live in now. The quote also serves as a predictor for where we are headed as the applications of artificial intelligence will expand greatly in the future. Such progress will enable medical and manufacturing breakthroughs as well as fuel leaps in human's creativity, explained Neil Jacobstein, Singularity University Artificial Intelligence & Robotics Co-Chair at FutureMed 2013 in Silicon Valley. The medical field, which for decades has been resistant to AI, is becoming increasingly allured by its possibilities, he noted.

Neil Jacobstein

AI is also being applied to the very small and the evolution of molecular manufacturing will be guided by AI algorithms, Jacobstein said. “What we are after is nanotechnology defined by capability—not size,” he explained. In particular, AI could be used for atomically precise fabrication, which enables macro-size objects to be built with extraordinary properties. “In your lifetimes, you are going to see this revolution happen,” Jacobstein said. The synthesis of AI and molecular manufacturing will yield such as single-molecular switches and biological machines with unique properties. The field of synthetic biology also draws on constraint-based AI techniques, which originated two decades ago.

Consumer AI systems like Siri and Google’s answer to it, Google Now, are the tip of the iceberg of what is possible with artificial intelligence. “Both of these are in the very early stages and they are going to get much better,” Jacobstein said. 

No recent discussion of artificial intelligence would be complete without mentioning IBM’s Watson, which famously trumped human opponents in Jeopardy! in 2011. “Watson is not just playing games,” Jacobstein added, pointing to IBM’s collaboration with the Memorial Sloan Kettering Cancer Center to create a decision support tool for cancer diagnosis.

IBM is also working on cognitive computing platform. The architecture of the platform is substantially different than traditional electronics. Spearheaded by Dharmendra Modra, project leader at the IBM Cognitive Computing Project, the project brings together a number of disparate technological disciplines. “[The platform] is a learning architecture. A fundamental difference in kind—not just speed,” Jacobstein explained.

“Eventually we are going to be able to build an artificial neocortex,” Jacobstein predicted. 

The clip below provides background information on Modra's cognitive computing research:

Jacostein's talk also touched on numerous other promising AI applications. To name but a few: Kaggle is a platform that aims to make a sport out of data science, which Merck has used to promising molecules for drug discovery; Archimedes simulates clinical trials; MEDgle is a promising health analytics platform; Practice Fusion is integrating clinical decision support into its EMR systems.

The rate of progress in this field is set to proceed at a rapid clip, Jacobstein predicted, as the systems are becoming ever more powerful as well as easier to program.

Brian Buntz is the editor-in-chief of MPMN. Follow him on Twitter at @brian_buntz.