Rocky Nevin

CEO, DataSea, Inc.

Human-friendly Computing through Natural Language Inferencing
16 minutes, 7.7mb, recorded 2008-03-14
Rocky Nevin

Rocky Nevin, CEO of Datasea gives a talk about the current state of human-computer interaction and what we need to do to improve this field and make it really usable. He argues that today's procedural and 'hieroglyph' based approach is not enough. Asking a question to a computer and having it relay an answer back to you that makes sense isn't something that sounds hard but is surprisingly a very complex task for a computer to perform.

Nevin proposes that instead of carefully structuring code to do what we want, we should instead follow nature's example, and have the data and the actions be self-organizing in order to create these complex data structures and algorithm to serve our human-computer interaction needs.


Rocky Nevin earned his Ph.D. in Biophysics at the University of California at Berkeley exploring the connectivity and structure of sensory neurons and identifiable interneurons in complex behavior. He worked at IBM's Palo Alto Scientific Center, and later headed the Computer Integrated Manufacturing group at MicroUnity Systems Engineering, where he wrote the visualization and control software for its $100 million wafer fabrication plant. Using his thesis work as inspiration, he invented a new approach to computing and network modeling, one that more closely mimics human-like inferencing abilities to derive answers from a fusion of data sets.

Resources

This free podcast is from our Emerging Communications series.

For The Conversations Network:

  • Post-production audio engineer: Robb Lepper
  • Website editor: Jimmy Karlsson
  • Series producer: Sathyaish Chakravarthy