"...and more..":
Want to confirm whether the names Bill H and Ken S ring as many bells in neural networking circles as the name Lowe does among mountain climbers?
Start with corporate bio data:
areteindustries.com
Next, who's got a new book out from MIT Press? Hint: it's a reference book for "for speech scientists... and engineers concerned with speech processing applications." Hint two: the sort of thing that lets Captain Kirk tell his computer what to do. Give up? It's Ken S.
mitpress.mit.edu
In this rarified circle, whose name gets used with such familiarity that it doesn't even get a footnote or require an entry in the bibliography? It's Hutchison! (Sort of like "Einstein? Is that Fred Einstein?") Check 3/4 of the way down for the following text: "With the development of the backpropagation algorithm, connectionist models with hidden layers could learn. This eliminated the
limitations of connectionist models and mooted Minsky and Papert's argument. Intiguingly, as Hutchison notes, the requirement of a hidden layer in order to make a connectionist network entirely computationally general only holds for feedforward architectures...'
Finally, at the May 2001 ABA (American Behavioralists?) conference, a symposium entitled entitled "Verbal Behavior and Technology: How Far Can It Go?" will include presentations by Bill Hutchison and about him! The following is from
http://#######.csustan.edu/pipermail/verbalbeh/2000-November/000093.html
****text follows***
FIRST PRESENTATION INFORMATION:
Presentation Title: An Operant Computer Model that Learns Verbal Behavior
Abstract: This paper describes an operant computer model and has been described at many previous ABA and SQAB conferences. The behavior analysts in Applied Behavior Systems are developing a software system to teach verbal behavior to developmentally delayed children. A major
long-term application of that software, however, is also to use it to teach the operant computer model the same verbal behavior as young children. Such
a project will not only serve to prove the sufficiency of
Skinner's analysis of verbal behavior, but will also have major practical implications. This first paper in the symposium will describe the overall approach and plan for this project.
First Author Information
William R. Hutchison
Applied Behavior Systems LLC
SECOND PRESENTATION:
Presentation Title: How Machines Learn to Recognize Speech
Abstract: More and more computer applications are designed with the ability to process human speech. What are the different methods by which machines can learn to recognize speech? How does our behavior-analytic approach differ from other
methods? What is the current state of technology for speech recognition? What are the potentials for speech recognition in the future and how are we likely to get there. My talk will address and attempt to answer all of these questions.
First Author Information
Matthew J. Morris
Applied Behavior Systems LLC
THIRD PRESENTATION:
Presentation Title: Intelligent Agents and Verbal Behavior
Abstract: The inclusion of "intelligent agents" in software products allows a type of user interaction that increasingly looks analogous to verbal behavior.
"Intelligent" agents come in many different varieties, but there is at least one type (Bill Hutchison's 7G) that has the ability to learn and acquire relationships that are verbal analogs. Such an agent is at the heart of the SpeechTeach product, and uses its own ability to acquire "verbal behavior" to guide the tutorial progress of human children. Is the response of seeing
a picture of a cup on the screen and saying "cup" into a
microphone a tact, if voice recognition software parses the response and an intelligent agent delivers conditioned reinforcement? Is it possible to train minds with a computer? How about autoclitic frames (grammar?) Data from the SpeechTeach project will be presented to suggest answers to these questions.
First Author Information
Ken Stephens
Applied Behavior Systems
*****end of text****
Being able to talk to your computer will be a kick. might be even better to tell about how you invested in the company that made it possible.
Just food for thought,
Caradoc
Regards,
John :-) |
