Shortly after I graduated from MIT in the mid-1960s, the Cambridge/Boston geek community was intrigued by the announcement of a computer program called Eliza, developed by MIT computer science professor Joseph Weizenbaum. I never had the pleasure of taking any courses from Professor Weizenbaum while I was in college, but I was fascinated by the idea of a computer program that did such a good job of mimicking the behavior of psychiatrists — in what I am told is a Rogerian style — that many lay people, and perhaps even a few computer-literate folks, thought that they were interacting with a computer program that really was intelligent.Stories abounded at the time of computer geeks who deliberately left the Eliza program running on their computer terminal when they went home from the office, in the hope that their co-workers might get fooled into interacting with the program.

Eliza would begin, back in 1966 as well as in today’s web-based version, by displaying the following message:

Hello. I am ELIZA. How can I help you?

If you responded by typing something like,

“My mother hates me,”

(as I did just now, with the Web-based version) it would then ask,

Who else in your family hates you?

And if you said,

Nobody. Just my mother.

Eliza would respond with the message:

Surely not everyone?

You have to admit it: it’s a little creepy. But you can also see what’s going on: Eliza looks for certain key words and phrases, and regurgitates your answer with a follow-on question. Once the basic algorithm has been articulated in this fashion, you tend to shrug — especially if you have any ability to program computers — and say, “Of course!” And I well remember Professor Weizenbaum’s prophetic comment, in a January 1966 Communications of the ACM article entitled “ELIZA — A Computer Program For the Study of Natural Language Communication Between Man and Machine,” that “to explain is to explain away”:

It is said that to explain is to explain away. This maxim is nowhere so well fulfilled as in the area of computer programming, especially in what is called heuristic programming and artificial intelligence. For in those realms machines are made to behave in wondrous ways, often sufficient to dazzle even the most experienced observer. But once a particular program is unmasked, once its inner workings are explained in language sufficiently plain to induce understanding, its magic crumbles away; it stands revealed as a mere collection of procedures, each quite comprehensible. The observer says to himself “I could have written that”. With that thought he moves the program in question from the shelf marked “intelligent” to that reserved for curios, fit to be discussed only with people less enlightened than he.

The object of this paper is to cause just such a reevaluation of the program about to be “explained”. Few programs ever needed it more.

Weizenbaum, as you may have gathered from this brief excerpt, was a humble man, and had no illusions about the intrinsic novelty of the program he had created. I had a chance to see this, first-hand, in early 1966 when I worked as a programmer at Digital Equipment Corp. (DEC). I can’t remember the details of how we did it, but (a) we got a version of Eliza running on our R&D DEC PDP-6 time-sharing system, (b) we dialed up the Project MAC time-sharing system at MIT with a low-speed communication link, and got a copy of Eliza running on that machine, and (c) we got the two copies of Eliza talking to each other. But the problem, as you can tell from the brief exchanges cited above, is that Eliza doesn’t want to (and is utterly incapable of) talking about itself: it wants to let you talk about yourself. So the exchange between the two Elizas degenerated very quickly into exchanges of, “Thanks, but let’s talk about you,” and “No, no, let’s talk about you.

It was hilarious, but also somewhat deflating to those of us who had studied under the likes of Marvin Minsky, and thought that AI would someday accomplish what Steven Spielberg envisioned in his 2001 movie, Artificial Intelligence.Much of Weizenbaum’s subsequent career was devoted to challenging and critiquing some of the lofty — and occasionally exaggerated — claims and predictions about the future of artificial intelligence. Some ten years after he developed Eliza, Weizenbaum published what I have long thought was one of the seminal books in the computer field: Computer Power and Human Reason: from judgment to calculation. I’m delighted to see that (used) copies are still available today from Amazon.

So why am I mentioning all of this? Simply because my wife — who has heard my stories about the DEC/MIT exploits with Eliza enough times to groan every time I repeat it, but who had no idea who Joseph Weizenbaum was — tore out a page from the obituary section of the New York Times yesterday, and brought it into my office.”Ever heard of this guy?” she asked. The obituary was titled, “Joseph Weizenbaum, Famed Programmer, is Dead at 85.”

Well, god bless you, Joseph Weizenbaum. I hope you and Eliza have been ushered through the pearly gates, and are enjoying the beginning of a peaceful eternity together.