Roger Schank
Artificial Intelligence

I worked for 25 years in Artificial Intelligence.

In the 60’s I was concerned with getting computers to understand English. In the 70’s I worked on getting computers to have a model of the world based upon knowledge that people have and use that knowledge to comprehend the world around them. In the 80’s I tried to make machines have sophisticated memories about events they had processed and be able to make predictions, recover from failed predictions and generally learn from experience. This became known as the field of Case-Based Reasoning.

Human beings are not rational planners. Decision-making is unconscious and rationalized consciously after a decision has been is made. Cases, known consciously and unconsciously, drive the human decision-making process. Any intelligent computer system that ignores how people make decisions will behave in a fashion that, while logical, may miss the forest for the trees. Real decision making relies on the power of a complex, well-indexed case base.


My new goal is not to build intelligent machines but rather to build an intelligently organized and indexed knowledge base that is easily accessible and gets smarter over time.


It is still a kind of AI, but the goal is more modest -- to make computers actually know something important and to be able to provide that information at the time of need.

Quotation mark

You can build learning machines, and the learning machine could painstakingly try to learn, but how would it learn? It would have to read the New York Times every day. It would have to ask questions. Quotation mark

--Roger Schank, Information Is Surprises
from The Third Culture: Beyond the Scientific Revolution

Gears
The Reminding Machine

The plan starts with a massive attempt at story collection. Stories would then be indexed by hand. When enough stories are collected and indexed, a knowledge base of what smart people know about how to deal with complex situations would be available to all who need it.

Read The Reminding Machine by Roger Schank

Octopus
Octopus Incorporated

A large company is like an octopus: it should have a central processor which can absorb the experience of each arm and track the goals it is pursuing.

Learn more about Corporate Memory

spacer
Related Articles
bulletWhy we hate email

Well, you wonder, isn’t that nice, but how could an e-mail program be that smart? It would have to know my business inside and out. Right. Why shouldn’t your e-mail program be smart enough to know which e-mails still need to be acted upon, and which present you with a potential legal liability issue simply because you have opened them, and which need to go on to a to do list of some type to be checked on regularly?

Read Why We Hate Email by Roger Schank


bulletThe mind of the octopus

Image a beast with eight arms. Now ask yourself whether one arm knows what the other is doing. One hopes that the octopus has a mind which serves as a central processor which can absorb the experience of each arm and track the goals it is pursuing.

The Mind of the Octopus by Roger Schank


bulletSplendid Learning

Every time I think that e-learning couldn’t get any sillier, someone figures out how to take a really bad idea and promote it into an ideology. At first I didn’t get what anyone could be talking about when they used the phrase “blended learning.” Those awful programmed learning workbook training manuals that everyone hated? Put them online with some cute graphics and presto: e-learning. Yippee!!

Splendid Learning by Roger Schank

Iceberg
Information that finds you

If I am captain of a ship and the port I am about to enter has a port pilot who expects bribes and an accident history when the bribes are not sufficient, could my computer let me know about this? It could, but it just doesn’t. And if I am about to make a decision about an engine that turns out to have been made by others in similar situations with bad effects, couldn’t my computer tell me about this before I do it? It could.

Information That Finds You by Roger Schank

Home
News
Advice Column
Blog
Media / Publications
Biography
Personal
Contact
Education
Training
Learning Games
Speaking
Artificial Intelligence
Corporate Memory
Intelligent Enterprise Software
Innovation