Screenshot. Doug Lenat”We’ve been keeping a very low profile, mostly intentionally,” said Doug Lenat, President and CEO of Cycorp. “No outside investments, no debts. We don’t write very many articles or go to conferences, but for the first time, we’re close to having this be applicable enough that we want to talk to you.”IBM‘s Watson and Apple‘s Siri stirred up a hunger and awareness throughout the United States for something like a Star Trek computer that really worked — an artificially intelligent system that could receive instructions in plain, spoken language, make the appropriate inferences, and carry out its instructions without needing to have millions and millions of subroutines hard-coded into it.
As we’ve established, that stuff is very hard. But Cycorp’s goal is to codify general human knowledge and common sense so that computers might make use of it.
Cycorp charged itself with figuring out the tens of millions of pieces of data we rely on as humans — the knowledge that helps us understand the world — and to represent them in a formal way that machines can use to reason. The company’s been working continuously since 1984 and next month marks its 30th anniversary.
“Many of the people are still here from 30 years ago — Mary Shepherd and I started [Cycorp] in August of 1984 and we’re both still working on it,” Lenat said. “It’s the most important project one could work on, which is why this is what we’re doing. It will amplify human intelligence.”
It’s only a slight stretch to say Cycorp is building a brain out of software, and they’re doing it from scratch.
“Any time you look at any kind of real life piece of text or utterance that one human wrote or said to another human, it’s filled with analogies, modal logic, belief, expectation, fear, nested modals, lots of variables and quantifiers,” Lenat said. “Everyone else is looking for a free-lunch way to finesse that. Shallow chatbots show a veneer of intelligence or statistical learning from large amounts of data. Amazon and Netflix recommend books and movies very well without understanding in any way what they’re doing or why someone might like something.
“It’s the difference between someone who understands what they’re doing and someone going through the motions of performing something.”
Cycorp’s product, Cyc, isn’t “programmed” in the conventional sense. It’s much more accurate to say it’s being “taught.” Lenat told us that most people think of computer programs as “procedural, [like] a flowchart,” but building Cyc is “much more like educating a child.”
“We’re using a consistent language to build a model of the world,” he said.
This means Cyc can see “the white space rather than the black space in what everyone reads and writes to each other.” An author might explicitly choose certain words and sentences as he’s writing, but in between the sentences are all sorts of things you expect the reader to infer; Cyc aims to make these inferences.
Consider the sentence, “John Smith robbed First National Bank and was sentenced to thirty years in prison.” It leaves out the details surrounding his being caught, arrested, put on trial, and found guilty. A human would never actually go through all that detail because it’s alternately boring, confusing, or insulting. You can safely assume other people know what you’re talking about. It’s like pronoun use — he, she, it — one assumes people can figure out the referent. This stuff is very hard for computers to understand and get right, but Cyc does both.
“If computers were human,” Lenat told us, “they’d present themselves as autistic, schizophrenic, or otherwise brittle. It would be unwise or dangerous for that person to take care of children and cook meals, but it’s on the horizon for home robots. That’s like saying, ‘We have an important job to do, but we’re going to hire dogs and cats to do it.’”
If you consider the world’s current and imagined robots, it’s hard to imagine them not benefitting from Cyc-endowed abilities that grant them a more human-like understanding of the world.
Just like computers with operating systems, we might one day install Cyc on a home robot to make it incredibly knowledgable and useful to us. And because Cycorp started from zero and was built up with a knowledge of nearly everything, it could be used for a wide variety of applications. It’s already being used to teach math to sixth graders.
Cyc can pretend to be a confused sixth grader, and the user’s role is to help the AI agent understand and learn sixth grade math. There’s an emotional investment, a need to think about it, and so on. Our program of course understands the math, but is simply listening to what students say and diagnosing their confusion. It figures out what behavior can it carry out that would be most useful to help them understand things. It’s a possibility to revolutionize sixth grade math, but also other grade levels and subjects. There’s no reason couldn’t be used in common core curriculum as well.
We asked Lenat what famed author and thinker Douglas Hofstadter might think of Cyc:
[Hofstadter] might know what needs to be done for things to be intelligent, but it has taken someone, unfortunately me, the decades of time to drag that mattress out of the road so we can do the work. It’s not done by any means, but it’s useful.