The Seminar on Dangerous Ideas

John Mallery

Where Government Science Policy Matters:

Two National Initiatives in Fundamental CS & AI


1 p.m. Wednesday, March 19, 2003



This talk will discuss the need for publicly-supported, large-scale efforts to develop:

  1. a new generation of high-productivity hardware/software platforms based on introspection and self-adaptation; and
  2. natural language systems for text understanding with broad but shallow semantic coverage.
Both efforts promise radical, paradigm-shifting advances in the ways we program computers and manage vast quantities of text, with highly-significant economic impacts. However, they are too risky for industry and too large or long-term for current funding vehicles. We will argue that these objectives are technically, financially, and politically within reach of a forward-looking science policy.