From ON Magazine, Number 4, 2009

Technology Predictions: Intelligence & Brute Force

By Andrew Odlyzko

Technology prediction is inherently hard. And it is even harder to predict how society will react to a new product or service.

Potential customers may sneer at a new technology, as happened a decade ago with application service providers (ASPs). Or they may embrace it, as seems to be happening with today’s incarnation of ASPs, cloud computing. (Of course, it is still too early to tell if what we see is truly an enthusiastic embrace, or simply hype generated to stimulate an enthusiastic embrace.) The presence of complicated feedback loops-hype can inspire creation of new applications, which will make a service more attractive and persuade people to try it-makes the task of prediction even harder. So it is no wonder that “progress by mistake” is not just frequent, but almost a rule.

An unexpected killer app

Technology can surprise on the upside as well as the downside. E-mail, which was specifically excluded from the design criteria for the ARPANET, became the “killer app” of that network as well as its descendant, the Internet. Who could have known at the time that the computer mouse, demonstrated by Doug Engelbart more than 40 years ago, would today still be the key device for human-computer interaction? And the World Wide Web, now 20 years old, spread slowly for several years, until the release of the Mosaic browser made it widely accessible-and then it caught fire. But even then, in the first few years, there was considerable speculation that even better tools for accessing information over the Internet might emerge.

What can we conclude from the long history of failed technology predictions? Wide experimentation is certainly called for, as well as maximizing the flexibility of new technologies, in order to accommodate demands that one did not foresee initially. One should not count on serendipity, but be prepared for it. And, of course, we should ride the technology curve, taking advantage of Moore’s Law and similar laws that provide predictable progress in information technologies, at rates that vary from field to field.

Under- and overestimating

Aside from the widely accepted principles above, there are a few other patterns that one can discern in the history of predictions about technology. Thus, although general technology forecasting is unreliable, some predictions have proven correct over an extended period of time. A fairly persistent pattern is the underestimation of the continuing increases in processing power, storage capacities, and communication bandwidth, and overestimation of the extent to which computers can be made to reason like people.

A striking example of this dichotomy is provided by J. C. R. Licklider’s book Libraries of the Future, published in 1965. Licklider has the best claim of anybody to be called the “grandfather of the Internet,” as he was the first one to point to computers as being primarily communication devices, not just computing ones, and he set up the program that led to the creation of the ARPANET. In his book, he made many predictions. Some, about development of computer networks, and about digital libraries becoming feasible around the year 2000, are among the finest examples of futurology. But those were based primarily on extrapolations from basic technology trends. Many of his forecasts were wrong, in particular those based on expectations that computers would acquire intelligence.

A similar pattern appears in other areas: Speech recognition has made great strides, primarily by exploiting more powerful technology to do massive pattern matching, rather than by the methods pursued in the 1960s of trying to get computers to understand human speech. Language translation followed the same pattern. And so did chess. The best computer chess programs can handily beat the best human players today, but not by imitating human thought processes. (That presents us with a mystery: Why are there no contests involving pairings of people and computers on each side?)

Human, not artificial, intelligence

With the Web, too, brute force has triumphed, although that brute force is directed by human intelligence, in the form of clever algorithms. (Clever algorithms were also needed for the advances in speech recognition, language translation, and chess.) The popularity of the Web obtained a substantial boost from the appearance of AltaVista, the first popular search engine. AltaVista’s breakthrough, later improved on by Google, was to demonstrate that with sufficient computing, storage, and communications resources, one could do effective, automated crawling and indexing. But AltaVista’s managers, for what seemed to be good business reasons at the time, made the misstep of switching their focus to making AltaVista a portal, and thus facilitated Google’s rise to dominance. Google succeeds largely through use of massive resources, with direction from clever methods, but not ones drawn from conventional AI.

The Web is evolving rapidly. And there are hopes for major breakthroughs based on computer understanding of the growing volume of digitized data. Yet, if we go by historical precedents, such hopes will be disappointed. Computing, storage, and communications are all progressing rapidly, even if somewhat less rapidly than they did a decade ago. Hence, it is most reasonable to expect the incremental improvements they will provide (together with improvements in standard data mining, visualization, databases, and related algorithms) will be the main contributors to the Web’s evolution.

Andrew Odlyzko is professor of Mathematics at the University of Minnesota. Before founding the Digital Technology Center at the University of Minnesota, he had a long career in research and research management at Bell Labs and AT&T Labs. He is widely known for an early debunking of the myth of Internet traffic doubling every three or four months and for demonstrating that connectivity has traditionally mattered much more to society than content has.

From ON Magazine, Number 4, 2009

From Ethernet to Enernet:

Bob Metcalfe on Standards, Serendipity, and Stubs

Bob Metcalfe has been involved—as a direct catalyst or a prominent observer—in a number of key milestones spanning the evolution of the IT industry: the birth of the Internet, the invention of Ethernet and local area networks, and the rapid adoption of the World Wide Web as the platform for linking information and people. Today, as a partner in Polaris Ventures, he invests in clean, low-cost energy solutions. Gil Press spoke with him at his office in Waltham, Massachusetts.

In the 20 years since the invention of the World Wide Web, what has surprised you most?

Bob Metcalfe: Tim Berners-Lee invented the URL, HTTP, and HTML standards. None of them is particularly impressive; so many high-tech people have found them to be in some way deficient. But Tim came up with three adequate standards that, when used together, ignited the explosive growth of the Web. The power of good standards is they leave you with no options. As we used to say about Ethernet, “anything which is not prohibited is mandatory.”

Think about that. We designed some plumbing at the lower levels of the hierarchy, and 17 years later Tim comes up with the World Wide Web, which Ethernet and TCP/IP carried just fine. That’s the surprise. What this has demonstrated is the efficacy of the layered architecture of the Internet. The Web demonstrates how powerful that is, both by being layered on top of things that were invented 17 years before, and by giving rise to amazing new functions in the following decades. Based on the artfulness of the design of the interfaces, you give rise to serendipity. In the design of his standards, Tim nailed down both expressive power and simplicity, allowing people to easily get started. It’s those three standards plus Mosaic, which added visual and graphical veneer, plus the evangelical verve of Tim Berners-Lee himself that were probably all pivotal in that early takeoff.

What has been a disappointment in the context of the World Wide Web—something you expected that didn’t pan out?

There’s no room for that. The Web has been so successful, there’s nothing disappointing about it. Tim Berners-Lee tells this joke, which I hasten to retell because it’s so good. He was introduced at a conference as the inventor of the World Wide Web. As often happens when someone is introduced that way, there are at least three people in the audience who want to fight about that, because they invented it or a friend of theirs invented it. Someone said, “You didn’t. You can’t have invented it. There’s just not enough time in the day for you to have typed in all that information.” That poor schlemiel completely missed the point that Tim didn’t create the World Wide Web. He created the mechanism by which many, many people could create the World Wide Web.

And the mechanism to connect not only information, as was his original vision, but now also connecting people with Web 2.0 applications. You recently started to use Twitter. Why?

I’m using Twitter because one of my partners, Mike Hirshland, accused me of having a generational problem. Young, hip people use social networks, and old farts don’t.

I used to be on the other side. I was helping to introduce LANs when there were all these old farts who thought that punch-cards were the way you did computing. The joke was that Ethernet would be adopted one funeral at a time. These people had to die. There was no way of changing their minds. So, I understand generational ossification. When my partner accused me of it, I decided to participate in this phenomenon so as to better understand it.

I’m beginning to find uses for Twitter. By tweeting my weight, I have involved my followers in a support group to help me lose weight. Knowing that I’m going to be tweeting my weight bears on my behavior. So there’s one application—the support group application.

My daughter is about to graduate from college, and she’s looking for a job. I have tweeted this fact, and I’m actually getting inquiries about my daughter from people who might want to see her résumé. So, that’s the job search application.

One of my hobbies is math puzzles, and I tweet them now and then. The most response I’ve ever gotten on Twitter was when I tweeted the fact that 111,111 squared equals 12,345,678,987,654,321. Then I noticed that in a lot of the re-tweets there was a tag that I was unaware of: number sign, nerd porn—this particular fact was considered nerd porn.

In the early 1990s, you argued in an InfoWorld column against wireless computing, advising readers to “wire up your homes and stay there.”

Let’s divide that into two discussions. I think that “wire up your home and stay there” is truer than ever. We’re at a time now where energy conservation is the next big thing, and one of the opportunities we have is the substitution of communication for transportation.

But you ask about one of my regrettable columns. In the early 1990s, there was a wireless bubble. There were a bunch of companies touting their modems and wireless mobility. But the modems didn’t work very well, and they were bigger than the computers. I said that wireless mobile PCs would be like porta-potties: Porta-potties are good and useful things, but as a general rule, the bathrooms that we use have pipes. So yes, there will be some wireless computers, but mostly we’ll use pipes because pipes have so much more capacity. I was right about it in 1993: That bubble burst, and all those mobile wireless companies went away.

I went on to say in my column that wireless computing will never be important. That’s where I went wrong, because along came Wi-Fi. When I was writing my column, I was often torn between being right and being interesting. Many columnists make the mistake of trying too hard to be interesting. You use various forms of hyperbole, like “There will never be anything like this.” Well, maybe there will be. But that’s not nearly as interesting as these hyperbolic comments.

I would like to point out that there is a figure of speech called hyperbole. It’s a Greek word. It’s been around for a long time, so I offer it in defense of some of my hyperbolic columns.

Around the same time, George Gilder coined the term “Metcalfe’s Law” to describe your idea that bigger networks are better. In the context of the layered architecture of the Internet, don’t you think one can apply “Metcalfe’s Law” to the layer of networking computers (the Internet), the layer of linking information (the Web), and finally, the layer of connecting people (Web 2.0)?

That’s a great point. I’d never thought of it that way. It wasn’t even called Metcalfe’s

Law when I first used it. It was a slide in a 3Com sales presentation. The goal of the slide was to give people a rationale for building bigger Ethernets. I drew a picture that put the three-node network below a critical-mass point, arguing that you needed to get to some higher number to achieve critical mass. That was the diagram that I gave to George Gilder in 1993. He called it “Metcalfe’s Law,” for which I’m grateful. The value of the network grows as N-squared—“N” being the number of machines connected to the network.

Networking PCs was a novel idea at the time. So what did you tell people they could do with the network?

When Ethernet first came out, our sales proposition was PFMTS—Print, File, Mail, Terminal, Stubs.

You may remember the IBM PC XT that came out in 1982. It had a 10-megabyte disk on it. No one could imagine what you’d do with 10 megabytes on your disk. So the idea that you might want to buy one PC with a 10-megabyte disk on it, and then share it over the LAN with cheaper diskless PCs, had traction. The same thinking applied to laser printers that were new and expensive. So share the printer, share the disk.

I like to think about it as shifting gears. The second gear was LAN e-mail. The big e-mail carriers of the time, like AOL and MCI, didn’t consider it e-mail, because my e-mails never left the building. But already in the early days of the Internet, we observed heavy e-mail traffic between Internet nodes within the same building. We called it “incestuous traffic”; it was surprising, even embarrassing, because Internet e-mail was originally conceived for long-distance communications.

T stood for terminal. There were all these minicomputers and mainframes still around in those days. You couldn’t throw them out, and all of them had dumb terminals. People would have a dumb terminal on their desk, and then they would have a PC on their desk. That didn’t make any sense. So you’d just write software that allowed your PC to be a dumb terminal so you could access the minicomputer or the mainframe.

Stubs were the APIs for accessing the underlying networking functionality, opening connections, closing connections, etc. This is the serendipity idea again. One such new idea came from Novell, which used the stubs to share access—not to a file, but to a database. This led to the first use of multi-user accounting systems that ran on top of the LAN. That’s how NetWare got its foothold and eventually blew past 3Com’s operating system.

You have been drawing interesting analogies from your experience with Ethernet and the Internet to what you invest in and speak about nowadays: Energy or what you call the Enernet.

I’ve been on this Internet speaking tour, a two-year book tour without a book. I felt I had a valuable contribution to make, looking at how we built the Internet and extracting the lessons from that, and then applying them to energy so we could solve energy problems sooner, better, faster. I think there are a lot of lessons to be learned, such as the value of decentralization, designing for abundance, or over-reliance on Washington.

I used to defend that analogy. I’ve now come full circle: I believe that energy is the Internet’s next killer app. We did mail, we did telephone, we did commerce, we did publishing, we did newspapers (we’re about to kill newspapers), and now we, the Internet, is going to solve energy. For example, they talk about a smart grid. A smart grid is a bunch of folks out there who want to build new networks to solve energy and they call it the smart grid. But instead of building an entirely new network, another silo, why not use the Internet as the control plane for the smart grid?

But it’s even deeper than that. That is, the very structure of the energy network—the actual transmission and distribution—needs to be like the Internet. So, it needs to be de-synchronized. Right now, to put energy on the grid, you need to synchronize frequency and phase to get onto it because it clicks with this 60-hertz centralized clock.

What the Internet did for communications was to take the clock out and put the clock in the packet so there wasn’t a big global ticking clock. I sent you the clock, and you were able to tick the bits at the rate that I told you to, so we de-synchronized the Net. We will end up de-synchronizing the power switching network and end up with power packet switching, like the Internet.

What’s more, the other thing we did to telecom is we added storage. The original Internet had no storage in it. Then these geniuses came up with the packet switch, with core memory for storing packets. Then we added disks to our computers. If you look at the Internet now, there is storage everywhere. So we’re going to “storify” the energy network. Right now, they have no place to put energy, so when they have excess energy, they don’t know what to do with it. Also, if renewables such as solar and wind are going to play any role, you need storage. I think storage is going to be big in this new energy network we have to build.

What will the Web look like or should look like in 20 years?

Thinking about the future of the Web or the Internet, I came up with a three-by-three matrix. On one axis are the three new kinds of traffic that the Web has to deal with: video, mobile, and embedded. On the other axis are the next three societal applications that the Web has to solve: energy, healthcare, and education. I look in each of those nine boxes for companies, opportunities, and progress.

Those three kinds of traffic have started arriving, but we have a long way to go. Video is brand new on the Internet, as far as I’m concerned. The mobile Internet has arrived, but it’s still happening. Then there’s embedded traffic. Ten billion microcontrollers are shipped every year, and only a tiny fraction of those are networked. Then there are the three new killer apps—energy, healthcare, and education—just sitting there. The Web has got to solve all three of those problems.

What will the Web look like in 20 or 30 years? It will be comfortable with those three new modes of traffic, and it will be solving those three problems.

]]> http://onlifeininformation.com/?feed=rss2&p=453 http://onlifeininformation.com/?p=447 http://onlifeininformation.com/?p=447#comments Tue, 26 Jan 2010 01:28:17 +0000 Gil Press http://onlifeininformation.com/?p=447

From ON Magazine, Number 4, 2009

From the Web of Documents to the Web of Data:

Tim Berners-Lee on the Future of his Invention

In 1989, while a fellow at CERN, the European Particle Physics Laboratory in Geneva, Switzerland, Tim Berners-Lee invented the World Wide Web. Today he is 3Com Founders Professor of Engineering at the Massachusetts Institute of Technology (MIT), where he serves as director of the World Wide Web Consortium (W3C), an international standards body dedicated to leading the Web to its full potential. Sir Tim is the author of Weaving the Web. Jason Rubin spoke with him at his office in Cambridge, Massachusetts.

Twenty years on, the World Wide Web has proven itself both ubiquitous and indispensible. Did you anticipate it would reach this status, and in this time frame?

Tim Berners-Lee: I think while it’s very tempting for us to look at the Web and say, “Well, here it is, and this is what it is,” it has, of course, been constantly growing and changing-and it will continue to do so. So to think of this as a static “This is how the Web is” sort of thing is, I think, unwise. In fact, it’s changed in the last few years faster than it changed before, and it’s crazy for us to imagine this acceleration will suddenly stop. So yes, the 20-year point goes by in a flash, but we should realize that, and we are constantly changing it, and it’s very important that we do so.

I believe that 20 years from now, people will look back at where we are today as being a time when the Web of documents was fairly well established, such that if someone wanted to find a document, there’s a pretty good chance it could be found on the Web. The Web of data, though, which we call the Semantic Web, would be seen as just starting to take off. We have the standards but still just a small community of true believers who recognize the value of putting data on the Web for people to share and mash up and use at will. And there are other aspects of the online world that are still fairly “pre-Web.” Social networking sites, for example, are still siloed; you can’t share your information from one site with a contact on another site. Hopefully, in a few years’ time, we’ll see that quite large category of social information truly Web-ized, rather than being held in individual lockdown applications.

You mentioned a “small community” of people who see the value of the Semantic Web. Is that a repeat occurrence of the struggle 20 years ago to get people to understand the scope and potential impact of the World Wide Web?

It’s remarkably similar. It’s very funny. You’d think that once people had seen the effect of Web-izing documents to produce the World Wide Web, doing likewise with their data would seem the next logical step. But for one thing, the Web was a paradigm shift. A paradigm shift is when you don’t have in your vocabulary the concepts and the ideas with which to understand the new world. Today, the idea that a web link could connect to a document that originates anywhere on the planet is completely second nature, but back then it took a very strong imagination for somebody to understand it.

Now, with data, almost all the data you come across is locked in a database. The idea that you could access and combine data anywhere in the world and immediately make it part of your spreadsheet is another paradigm shift. It’s difficult to get people to buy into it. But in the same way as before, those who do get it become tremendously fired up. Once somebody has realized what it would be like to have linked data across the world, then they become very enthusiastic, and so we now have this corps of people in many countries all working together to make it happen.

Do you see the Semantic Web as enabling greater collaboration between and among parties, as opposed to the point-to-point or point-to-many communication that seems more prevalent in the current Web?

The original web browser was a browser editor and it was supposed to be a collaborative tool, but it only ran on the NeXT workstation on which it was developed. However, the idea that the Web should be a collaborative place has always been a very important goal for me. I think harnessing the creative energy of people is really important. When you get people who are trying to solve big problems like cure AIDS, fight cancer, and understand Alzheimer’s disease, there are a huge number of people involved, all of them with half-formed ideas in their minds. How do we get them communicating so that the half of an idea in one person’s head will connect with half of an idea in somebody else’s head, and they’ll come up with the solution?

That’s been a goal for the Web of documents, and it’s certainly a goal for the Web of data, where different pieces of data can be used for all kinds of different things. For example, a genomist may suspect that a particular protein is connected to a certain syndrome in a cell line, search for and find data relating to each area, and then suddenly put together the different strains of data and discover something new. And this is something he can do with the owners of the respective pieces of data, who might never have found each other or known that their data was connected. So the Web of data will absolutely lead to greater collaboration.

Is your vision of the Semantic Web one in which data is freely available, or are there access rights attached to it?

A lot of information is already public, so one of the simple things to do in building the new Web of data is to start with that information. And recently, I’ve been working with both the U.K. government and the U.S. government in trying not only to get more information on the Web, but also to make it linked data. But it’s also very important that systems are aware of the social aspects of data. And it’s not just access control, because an authorized user can still use the right data for the wrong purpose. So we need to focus on what are the purposes for accessing different kinds of data, and for that we’ve been looking at accountable systems.

Accountable systems are aware of the appropriate use of data, and they allow you to make sure that certain kinds of information that you are comfortable sharing with people in a social context, for example, are not able to be accessed and considered by people looking to hire you. For example, I have a GPS trail that I took on vacation. Certainly, I want to give it to my friends and my family, but I don’t necessarily wish to license people I don’t know who are curious about me and my work and let them see where I’ve been. Companies may want to do the same thing. They might say, “We’re going to give you access to certain product information because you’re part of our supply chain and you can use it to fine-tune your manufacturing schedule to meet our demand. However, we do not license you to use it to give to our competition to modify their pricing.”

You need to be able to ask the system to show you just the data that you can use for a given task, because how you wish to use it will be the difference in whether you can use it. So we need systems for recording what the appropriate use of data is, and we need systems for helping people use data in an appropriate way so they can meet an ethical standard.

Ultimately, what is one of the most significant things the Semantic Web will enable?

One thing I think we’ll be able to do is to write intelligent programs that run across the Web of data looking for patterns when something went wrong-like when a company failed, or when a product turned out to be dangerous, or when an ecological catastrophe happened. We can then identify patterns in a broad range of data types that resulted in something serious happening, and that will allow us to identify when these patterns recur, and we’ll be better able to prepare for or prevent the situation.

I think when we have a lot of data available on the Web about the world, including social data, ecological data, meteorological data, and financial data, we’ll be able to make much better models. It’s been quite evident over the last year, for example, that we have a really bad grasp of the financial system. Part of the reason for that might be that we have insufficient data from which to draw conclusions, or that the experts are too selective in which data they use. The more data we have, the more accurate our models will be.

After 20 years, what about the Web-either its current or future capabilities-excites you the most?

One of the things that gets me the most excited are the mash-ups, where there’s one market of people providing data and there’s a second layer of people mashing up the data, picking from a rich variety of data sources to create a useful new application or service. A classic example of a mash-up is when I find a seminar I want to go to, and the web page has information about the sponsor, the presenter, the topic, and the logistics. I have to write all that down on the back of an envelope and then go and put it in my address book; I have to put it in my calendar; I have to enter the address in my GPS-basically, I have to copy this information into every device I use to manage my life, which is inefficient and time-consuming. This is because there is no common format for this data to become integrated into my devices.

Now, the vision of Semantic Web is that the seminar’s web page has information pointed at data about the event. So I just tell my computer I’m going to be attending that seminar and then, automatically, there is a calendar that shows things that I’m attending. And automatically, an address book I define as having in it the people who have given seminars that I’ve attended within the last six months appears, with a link to the presenter’s public profile. And automatically, my PDA starts pointing towards somewhere I need to be at an appropriate time to get me there. All I need to do is say, “I’m going to that seminar,” and then the rest should follow.

The Web is such a mélange of useful, noble content and stuff that runs the gamut from the mundane to the grotesque. Do you think humanity is using this incredible invention of yours appropriately?

Yes. The Web, after all, is just a tool. It’s a powerful one, and it reconfigures what we can do, but it’s just a tool, a piece of white paper, if you will. So what you see on it reflects humanity-or at least the 20 percent of humanity that currently has access to the Web.

As a standards body, the W3C is not interested in policing the Web or in censoring content, nor should we be. No one owns the World Wide Web, no one has a copyright for it, and no one collects royalties from it. It belongs to humanity, and when it comes to humanity, I’m tremendously optimistic. After 20 years, I’m still very excited and extremely hopeful.

Google is now officially the evil empire for all other big tech players, as Microsoft used to be. Every day, there is yet another news flash on how someone is standing up to the Big G, with Apple recently leading the charge.

As Nick Carr observed, followed by WSJ’s Holman Jenkins, Google has gone to war with China because the issue at stake, the secured privacy of our email accounts (and other personal information) on the Web, is crucial to Google’s business. But today I realized that Google is also reacting to the increasing pressure from competitors that promote not only competing products, but also the perception of Google as the new evil empire. The best PR strategy for a counter-attack is to align yourself with a “don’t be evil” issue which is at the heart of the Web, at least as it is perceived by Americans: The free flow of information. Here’s a couple of reactions that Google’s PR people expected or at least hoped for as a result of declaring war on China: GigaOm’s Liz Gannes finds China’s response “seriously offensive,” and WSJ’s L. Gordon Crovitz calls the altercation “a Barbary moment for the Information Age.”

I understood Google’s PR game today as a domestic competitive reaction when I turned on my Blackberry. I found a new icon on it, one I never downloaded on my own. Yes, you guessed, a Bing icon. Looks like the RIM folks are not far behind Apple in joining forces with the previous evil empire to kick the current holder of the title. Now, who wants to predict when will Apple become the new Evil Empire?

The next entry in ON magazine’s Web Timeline stars Katharine Hepburn.

“They can’t build a machine to do our job; there are too many cross-references in this place,” she tells her anxious colleagues in the corporate research department in the film Desk Set (1957). The cause for anxiety is EMERAC, an ominous-looking, room-size computer, and a “methods engineer” (Spencer Tracy) who was hired to “improve workman-hour relationship.” By the end of the film, Hepburn proves her point by winning, not only the engineer’s heart, but also a contest with the “Electronic Brain.”

The film, particularly the statement about cross-references (one of the themes of this timeline), has served as a source of inspiration and a rallying cry for librarians ever since. I learned about it when I joined in 1988 the corporate research department (which included a magnificent network of libraries) at a leading purveyor of “electronic brains,” Digital Equipment Corporation.

In 2002, Cheryl Knott Malone published in the IEEE Annals of the History of Computing an interesting article on Desk Set and its historical context. The context was the promise and the threat of the new technology, its perceived potential to mimic (and replace) the human brain. That meant not only being able to think and reason like human beings, but also to suffer from the similar memory lapses and failure of over-worked circuits.

While correctly describing Desk Set as “the lone film [at the time] to offer a vision of the future of computing outside the realm of science fiction,” she ignores the fictional elements of the computer’s portrayal in the movie. She even states that the script (and the play on which it was based), “was prophetic in its depiction of automated information retrieval by an electronic computer capable of storing the contents of an entire library, processing natural-language input, and prompting users to formulate more precise queries.” Only the first attribute became true: no computer, then and now, could converse with its users human-like, the way EMERAC does it in the movie.

By anthropomorphizing the computer (called “Emmy” by the engineers working with it), the movie puts the fears of the new technology to rest - it is fallible, after all, and it certainly cannot replace the librarians, only assist with mechanical, repetitive work, such as processing the payroll.

But the anxiety about the new technology did not go away. In surveying librarians’ attitudes at the time, Malone quotes Jesse Shera, Dean of the library school at Western Reserve University and one of the first proponents of the synergy between information science and library science. Shera was chief sponsor of a book (which Malone does not mention) summarizing a study of the “promise and threat” of computers in libraries, The Librarian and the Machine, written in the early 1960s (published 1965) by Paul Wasserman.  The book opens with the following declaration: “As if it were not already a problem enough for library administrators striving to respond to the myriad pressures and infinite complexities…. A newer and even more foreboding terror [my italics]… was beginning to emerge more clearly. I refer to the computer and its attendant supporting apparatus.” But, like Desk Set, his conclusion at the end of his year-long study is comforting: “Machines today can do much of man’s work more rapidly and efficiently; but they cannot do his intellectual work as well.”

Why was it - and still is - even conceivable that a computer can do man’s thinking, or more narrowly, replace a librarian?  Because of Giant Brains like Turing, von Neumann, and Wiener, and popularizers such as Edmund Berkeley.  The Special Issue of ON also includes a great article by Andrew Odlyzko (p. 48), showing how this misplaced and unfounded belief in the potential of computer technology to become equal or even to surpass human intelligence has led to failed predictions about the trajectory of computer technology: “A fairly persistent pattern is the underestimation of the continuing increases in processing power, storage capacities, and communication bandwidth, and overestimation of the extent to which computers can be made to reason like people.”