You may not know Joe Weizenbaum's name, but many people are familiar with the computer program he wrote more than 40 years ago, Eliza. Eliza mimics a Rogerian psychotherapist, picking up key words you type in and spitting them back in the form of questions:
You: "I feel anxious today." Eliza: "That's interesting. What are you anxious about?" Etc.
In some ways it was very simple minded and Joe himself considered it a parody of psychotherapy. But in other ways it struck a deep chord. It was one of the first computer programs to simulate a human conversation and to give the impression of a sentient being in the box. Joe told me once how stunned he was one day to come into his office to find his secretary having a session with Eliza. She asked for privacy. Disturbed by her failure to recognize this was just a machine and moreover one she knew her boss had programmed, he set to work on his groundbreaking book, Computer Power and Human Reason (1976), multiple drafts of which he circulated via a Department of Defense electronic network called ARPANet, the forerunner to the internet. I have a typescript of it somewhere, as well as an autographed copy of the published version.
It was a highly controversial book in the computer science community, at least that part of the community involved in what became known as Artificial Intelligence. Joe excoriated those who intimated there was any equivalence between Artificial Intelligence and Human Intelligence. He worried the failure to understand the difference between a programmed response and genuine intelligence was a slippery slope to the kind of seductive and silent acquiescence to immorality that drove his family from Nazi Germany in the 1930s. From his obituary in the New York Times:
"He raised questions about what kinds of relationships we want to have with machines very early," said Sherry Turkle, a professor in the program in science, technology and society at M.I.T. who taught courses with Mr. Weizenbaum on the social implications of technology.
Mr. Weizenbaum also believed that there were transcendent qualities in the human experience that could not be duplicated in interactions with machines. He described it in his book as "the wordless glance that a father and mother share over the bed of their sleeping child," Ms. Turkle said.
The book drove a wedge between Mr. Weizenbaum and other members of the artificial intelligence research community. In his later years he said he came to take pride in his self-described status as a "heretic," estranged from the insular community of elite computer researchers. (Obituary, New York Times)
When I was a young academic I had the tremendous good fortune to know an unusual group of public intellectuals: Phil Morrison, Noam Chomsky, Jerry Lettvin, Steve Gould, Ruth and George Wald, Dick Levins, Dick Lewontin -- and Joe Weizenbaum. Some of them I knew better than others and Joe was one I knew well. Joe and his wife Ruth came to dinner one night at our small apartment and Mrs. R., who is a fine cook, made lasagna. She was nervous about entertaining someone of his stature. About halfway through the meal, Joe asked casually if there were any eggs in the meal and Mrs. R. said, "no." Later he mentioned that Ruth was very allergic to egg whites, and it was then I saw Mrs. R.'s face go white as well, as she remembered there were indeed eggs in the lasagna. Ruth seemed fine and nothing happened, but Mrs. R. has not forgotten it, even over the space of more than 35 years.
When we saw Joe's obit in the newspaper, Mrs. R. told me she had a real soft spot in her heart for him because, although when she first met him he was already a famous public intellectual and she was still working as a lab tech (she later earned a doctorate), he never made her feel inferior or stupid. He was invariably kind and respectful of others, and his warmth was irresistible. I saw Joe fairly often in those days, but even more often after his long marriage broke up. We used to meet frequently for coffee at a local cafe and spent many hours just talking: science, politics, what was going on in the world. Widely read, sophisticated and very smart, he was an extremely moral person, with high principles not always easy to live up to. When he went off to live in Europe in his later years we lost touch and I had not seen him for some time.
So now he is gone, dead of cancer at the age of 85. I can't say anything to him any more. But I can say to the rest of you what I feel: it was a privilege to know this man.
- Log in to post comments
revere(s): My condolences on the loss of your friend. He sounds like a wonderful man.
"Mr. Weizenbaum also believed that there were transcendent qualities in the human experience that could not be duplicated in interactions with machines. He described it in his book as "the wordless glance that a father and mother share over the bed of their sleeping child," Ms. Turkle said."
I found this an interesting quote. It highlights one of the issues that keeps me an agnostic instead of atheist. Probably a chicken/egg thing, but I am intrigued by the way we have evolved to combine our inborn instinct along with rational thought and the capacity for empathy with others. It doesn't prove the existence of some Higher Power, but it seems to beg for some better way to understand how we humans have come to be the way we are.
The story of the dolphin guiding the lost whales back to deeper water kind of hints at the same thing, to me. So many living things just follow a rigid set of behaviors: build the hive, feed the queen, dig more chambers for eggs, make more baby sharks, for instance. Something in humans allows us to be more imaginative, allows us to do things which are not just self-serving in some evolutionary way. Will machines be able to "transcend" in that same way?
Of course, we humans still are terribly destructive, to ourselves and everything and everyone around us. Most often, it seems we would be better off to just live guided by instinct. It wouldn't be pretty, but I can't imagine it could be any worse than it is right now.
Thanks for sharing about your friend, revere.
revere I've been nice and I really do not want to upset this fine tribute to a good man but I can't let this pass.
Puuleeaasseee !! I cannot believe, well yes I can, it is up there in black and white. Wenchacha you really did not, cannot, possibly mean that you cannot imagine it being any worse, if humans did not evolve into being what we are today? I doubt Mr. Weizenbaum would feel the same.
pauls: First off, you've been fine. Complaining and arguing around here are encouraged and it's OK to do both. Since your point is addressed to wenchacha I probably shouldn't respond, except I will say that the question can be sharpened, better or worse for whom or what? From the standpoint of human (as a species) destruction of the environment and deleterious effects on other creatures we share the planet with, the technology with the worst downstream consequences (and from which all else flows) was probably domesticaing agriculture. I'm not sure what to make of that and nothing about what has happened or will happen is pre-ordained. We always have choices about what we do and how we do it.
revere - something else would have domesticated and farmed on a grand scale that we humans enjoy. Ants for example. They milk aphids, they grow fungus. Hell they even war with one another. So I'm just glad we humans got there first and that everything else is downstream from us. Ants give me problems enough around here come spring. Damn little buggers. I believe domesticating animals and farming grew out of the human species wanting to survive - just like the ants learning to milk aphids and grow fungus. While most folks are in awe of the ants ability to accomplish these tasks, we shrug our shoulders at the human capacity at that point in time to evolve so that we could domesticate animals and farm. We must have been a tad smarter than ants of course, but when it comes to survival being smarter is a good thing.
The manufacturing guys over at Evolving Excellence have another positive tribute from a slightly difference perspective.
http://www.evolvingexcellence.com/blog/2008/03/joseph-weizenba.html
Ken
pauls: From the standpoint of biological success, I think the Insecta are considerably more successful than humans -- which is why you have trouble with them.
Ken: Thanks. I'll pop over and have a look.
yes revere but they don't know that, so let us keep this among we humans ok?
Wenchacha says: "Of course, we humans still are terribly destructive, to ourselves and everything and everyone around us. Most often, it seems we would be better off to just live guided by instinct. It wouldn't be pretty, but I can't imagine it could be any worse than it is right now."
----------------------------------------------
Transcendence is easy to be misinterpreted as supernatural or having the connotation of superstition; in fact it depicts the depth of reality, perhaps by a new term- hypothetical realism.
I could imagine a conscientious person like Joseph who had experienced Nazi's collective insanity; Nazi German had not lack of technology and super-machines and yet led into a despair of civilization. Therefore, his reasoning of the limitation of human intelligence in material reality in fact was a kind of prophesy, unfortunately seen as heretic in the United States' mainstream.
It is a sign of warning that perhaps the United States is stepping into what Arnold J. Toynbee and Oswald Spengler's decline loop; the rejection of genuine religious quest in the society at large.
So, instinct has to combine with clear conscience. The history of 20th century provide us a reasoning of European civilization; its twilight of civilization by a true example and message of Joseph which would deepen our reflection on our endeavor and belief system. Hence, we have a better chance to avoid the tragedy happened in the last century.
Do we need to stay as agnostic stance? Maybe most of our life needs it. Nevertheless, the leap of faith by hypothetical realism; the experience of transcendence and revelation ( The revelation could be understood as a delayed processing of past experience, an extremely rapid scanning procedure and sorting process.) in religious dimension need constant practices and explorations.
Agnostic stance indeed is humble, it get rid of hubris. It is a pathway of faith. Please combine with hypothetical realism for deeper reasoning.
pauls: I guess I wasn't thinking about all the downstream consequences of human evolution. I still believe that many aspects of instinctual behavior would be preferable to chosen ones. Animals can be greedy or wasteful, but I wonder if that is just a supply/demand issue, for instance. Bears take bites of salmon during their run, but they find the berries when they need them.
Humans do have a great many wonderful qualities! Why is so much of human history about our rivalries with each other, our desire for superiority, our withholding of compassion for those not on "our side"? It is truly a mystery to me that, being capable of choosing to be altruistic, unselfish, creative, instead the choice is to be spiteful, greedy and destructive. Even when we know we are capable of choosing, rather than just reacting instinctively.
There are plenty of dogs who can behave but also get into mischief. Maybe they have greater powers of reasoning than we know of at this time, but they are still dogs. Humans, on the other hand, always or nearly always have the choice to work in harmony with their community/environment instead of causing mischief. It's like we choose not to "transcend." Why? Does the dose of adrenaline or testosterone or whatever outweigh the endorphin bath?
We are just capable of so many more good things than we have actually done, as a species. I know we would not all be able to agree what constitutes "good" among those things.
This is the kind of thing that causes me and billions of other humans over time to wonder: why are we here? Is there any explanation for our human condition of being able to envision perfection/utopia but being unable (so far) to achieve it? Which I suppose causes us to seek out or believe in some unseen force that controls what we cannot.
I hate not having a satisfactory answer.
I would also like to offer my condolances, but I do not pity you for your loss; I envy you for it. Coffee with Weizenbaum. The conversations I've had with ELIZA have been fun, but definitely no substitute for that.
revere, many thanks for drawing the attention of this still sort-of-young academic to this fine gentleman and scholar. I am sad for your loss but am equally envious that you and Mrs R had the pleasure of knowing the man himself.
Racter, Abel: Yes, I was most fortunate. Coffee with Joe was always a pleasure and I did it often. He was the real thing, a European intellectual of the Old School. and a Mensch.
My condolences Revere; it indeed sounds like you were lucky to know him.
(I feel one should mention severe food allergies when accepting a dinner invitation -- yikes!)
I don't know if I would have been as disturbed by the secretary using the therapist-mimicking program. One wouldn't necessarily have to ascribe intelligence or emotional presence to a machine to find it useful to have one's statements echoed back.
It's rather like how I see Tarot card readings: I don't think the cards know anything, and I don't believe there's any larger force moving them into place. But I have observed them act as a prod for people to rethink what has/is/might happen. The cards aren't creating the thoughts; they, the subject, and the person doing the reading (if it's someone else) are interpreting them and using them to explore.
Sure, there are people who believe Tarot cards have some mystical power. But people believe in all kinds of things, which also can manifest in destructive ways. Maybe the secretary believed the machine was talking to her. I can see the dangers in taking instructions from (supposed) AI, but as long as it was generating questions by repeating, rather than providing answers, it seems any potential harm could only be created by the subject.
Apologies if it is bad form to debate with the recently deceased. I just found it a thought-provoking incident.
Ambiguous antecedent alert:
The "they" should refer to the humans, not the cards.
caia: No, not at all bad form. Joe would have loved to engage you in conversation on the topic. I appreciate the condolences but in reality I hadn't seen him in years, since he moved back to Berlin. I have no sharp sadness, just good memories and appreciation for having had him as a friend those many years ago. They don't make them like that anymore -- not worse, but different.
caia: well once you start describing the secretary (or whoever is using the program) as a 'subject', then all bets are off. I prefer user.
I wasn't thinking subject in terms of study subject; I meant subject as in actor, as in subject/object. But use user if you prefer.
pro-Weizenbaum anti-Minsky then ?
I found this:
http://blog.whoiswho.de/stories/2447/