Joseph Weizenbaum is best known in the English-speaking world for his 1976 popular critique of artificial intelligence, Computer Power and Human Reason. His reputation in Europe continued to flourish, however, as he wrote and spoke for German-speaking audiences until his death in 2008. Islands in the Seeking Havens of Reason in a Programmed Society is an extended interview with Weizenbaum, originally published in German in 2006. Imaginitive, iconoclastic, and always insightful about the role of computing in society, this book is a great introduction to the thought of Joseph Weizenbaum as it has evolved over the decades.
Insightful intro to Weizenbaum, especially since its published much more recently than his 1976 work. Overall, a little vague here and there with arguments not being fleshed out, but it is a 'stream of consciousness' style interview with interesting ideas.
See below, some examples from Chapter 4, "Television and Internet", regarding the utility of the internet for information and answering 'good' questions.
"The “screen-reality” is an edited, produced truth, subject to specific aims. It is the truth according to those who selected and combined it. Today we can manipulate images, television images, so that we can no longer tell if they are real or not. We can never really be sure if what we are seeing on the television is reality or if it is simulated."
"Next to the television screen as Fountain of Truth there is now the computer screen. Both screens reinforce each other. The World Wide Web appears to contain everything that is important in the world. It delivers an unmanageably large data flood into the house; it’s astounding. However, the consequence is actually a flood of disinformation."
"Besides – and we must not overlook this – this gigantic data flood, just like a gigantic choice of television programs, is a selection made by the people who put it in the ‘net. The Internet is a big garbage dump – admittedly with some pearls in it, but you have to find them first."
"But the Internet does offer many people the possibility to get a large quantity and a broad diversity of information, people who for various reasons were closed out of that. In certain circumstances; we should be very careful with the word ‘information.’ Its use is not only inflationary; it is also applied falsely. The signals in a computer are not information, they are just signals. There is only one way to turn signals into information, namely, to interpret the signals. To do that you need to use a human brain, because interpretation naturally means intellectual work. Coming back to the garbage dump in the Internet: We mustn’t forget that the Internet has become mass media and it almost seems like a Natural Law that every mass medium produces up to 95% nonsense. We were just talking about television. It’s always the same: In the beginning there are hopes, wishes, visions connected to the mass media, be it radio or television. People were generally optimistic, even euphoric. People expect, for example, that the general level of education will improve."
"The Internet has reawakened the same hopes. For every question or problem today, there is something to be found in the Internet, in the World Wide Web. Picture the Library of Congress in Washington. It is a monumental building. Picture a family visiting this landmark. The parents tell the children, that here you can find almost the entire written wisdom of humanity. Maybe the little daughter or the little son thinks, “Then I’ll ask a question and wait for the answer.” But it’s not that simple. Also because of the volume of data and material that is accessible through the Internet, we have the illusion that we only have to ask a question and the right answer comes right out. Unfortunately, for a number of reasons, that isn’t true. The Internet doesn’t work like a vending machine, where I throw in a coin and get what I wished for."
"Naturally there are pearls in the Internet, but to find them, the user needs a certain competency. You have to choose a specialist field in which you already know your way around. You have to know enough, in any case, to be able to formulate a good question."
"For me a good question is comparable with a plan for an experiment, for example in physics. First you have to know a lot, and then establish a connection, and then construct a physics experiment. That’s how you ask Nature a question. Behind it stands the thought, or the plan of the experiment."
Bit expensive too (+$60 AUD) and best I could do to access immediately from the middle of nowhere was to purchase the ebook version direct from the publisher, convert to kindle format, connect my kindle to PC and drag into kindle docs folder.
The author was a child emigré from Nazi Germany to Detroit where he became a mathematician and computer scientist at Wayne State. Among his early projects he created Eliza, the "talk-therapy" computer program. He never stopped reflecting on that experience -- which was not so much a technical experience as a widely impactful social one -- or on computer science or "artificial intelligence," which he critiqued during his whole life -- most of it spent as a colleague of all the "AI Elite" at M.I.T.
Do not overlook Professor Weizenbaum's point of view while we experience this current AI bubble / AI-for-everything / AI moneystorm. This book is an extended interview done near the end of his life so a sweeping look at it all. Also very accessible.
Only negative: $50 is too much for such a slim book, so try to borrow it.