Robby Garner builds chat bots. He's a natural language programmer exploring artificial intelligence (AI), and is listed in the Guinness Book of World Records as having written the "most human" computer program. Twice, his program Albert One received the Loebner Prize, the first formal award and contest for subjecting AI creations to the Turing Test.

One of the first web chatterbots, named Max Headcold, was written by Garner in 1995, entertaining customers of the Fringe Ware online bookstore and collecting data about web chat behavior. Garner and Fringe Ware co-founder Paco Xander Nathan evolved Max into a program called JFRED, which took part in the world's largest online Turing test in 1998. Their JFRED program was perceived as human by 17% of the participants.

These days, his bots are starring in a play, opening in Austria in September and at the COIL festival in New York City this winter. "Hello Hi There" is digital performance/laptop theatre creation of American director Annie Dorsen, who uses the 1970s interview between philosopher Michel Foucault and linguist Noam Chomsky as the basis for a dialogue between two chat bots developed by Garner for the occasion.

We recently chatted up Robby (known to some by his old business name, Robitron) to see if he's a man or a bot. Here's what we found.

What does AI really mean, to you? Do you find that amateurs have a different idea of what AI should be?

People probably think of AI as GAI, which is general artificial intelligence. They think of the Terminator movies, or of Hal 9000 in 2001: A Space Odyssey. They think of S.A.R.A.H. the house computer when they think of AI.

To me AI means more specific kinds of intelligence that in a whole, might lead up to general artificial intelligence.

The more specific kinds of AI are like speech recognition, search engine technology, machine vision, remote vehicle navigation, things like that. AI has had success with specialized topics like these, but has failed to provide a form of generalized intelligence like humans possess.

You create chat bots. What's the relationship between these and AI?

A chat bot is related to an area of AI called natural language processing (NLP). Some approaches to NLP are computational linguistics, which is often derived from grammatical theories, and the area that I work in is conditioning, which is behaviorist, in particular computational behaviorism.

Do you think it's important that you, or we the people, invent fully autonomous artificial intelligences?

I think that ultimately that is a goal worth working towards, and some of my colleagues have made steps towards that. However, in natural language understanding, there is still a long way to go towards human capabilities and human standards of intelligence.

How do discoveries and theories in AI affect science as a whole? Or society as a whole?

The advances made in google.com, for instance, transform natural language queries into usable information. Wolfram alpha has a natural language interface. These kinds of advances, and those of everyday spelling checkers affect many people, and are just examples of specific kinds of advances made in the field.

OK, so what's Wolfram alpha exactly?

Wolfram alpha is a computational knowledge engine which is like a search engine that can do complex math and knows about formulas for unit conversion, calculus, physics, and things of that nature.

You're spun back or forward in a time machine. What date to you choose to visit, and why?

I'd go to 2030 and see where the space program has gotten to. By that time, chat bot technology should have advanced beyond the hollow shell technologies that exist now, and on towards truly thinking machines.