Understanding the machine

Last week’s VCU’s New Media Faculty-Staff Development Seminar took up two related but also quite distinct essays: Norbert Wiener’s “Men, Machines, and the World About” and J.C.R. Licklider’s “Man-Computer Symbiosis.” Aside from the regrettable (but understandable) androcentric language, both essays are forward-looking, yet in different ways. Each of them understands that human history moves in the direction of greater complexity, especially in the accelerating streams of technological innovation and invention. (Wiener wrote a whole book on the subject of invention, one well worth reading, though it was not published until years after his death.) Both writers write about machines, systems, and human-machine interaction. Both writers emphasize that the computer is a new kind of machine. Wiener writes of a “logical machine” with feedback loops, and Licklider emphasizes the “routinizable, clerical” capabilities of the computer. Although neither one uses the magical phrase “universal machine” that Alan Turing uses, they both seem to understand that a difference in degree (speed, memory) can mean a difference in kind. Wiener also writes of “the machine whose taping [i.e., programming] is continually being modified by experience” and concludes that this kind of a machine “can, in some sense, learn.” Such machine learning, and research into its possibilities, is going on all around us today, and that pace too is accelerating. (Google Translate is but one example. Notice that it keeps getting better?)

Part of the experience computers learn from, of course, is our experience–that is, computers can be made and programmed so that they adapt to (learn from) our uses of them. It was hard to see this happening in the pre-Internet era. We could customize various things in DOS, and on the Macintosh, and on Windows (yes, even on Windows), but we didn’t have the feeling of the computer adapting to our uses. For that phenomenon to become truly visible, we needed the World Wide Web and cloud computing. (If you see an unidiomatic translation in Google Translate, click on the word, and Google Translate gives you the opportunity to teach it something.) The computer that learns from us most visibly is the computer formed of the decentralized, open, ubiquitous Internet, as that medium is harnessed by various entities. The most powerful application ever deployed on the Internet, the platform that enabled the macro-computer of the Internet to become visible and self-stimulating, is the World Wide Web.

Which leads me to my point, one already made more elegantly by Michael Wesch (see “The Machine is Us/ing Us“), Kevin Kelly, and Jon Udell, among many others. As we publish to the Web, purposefully and variously and creatively, we also make the Web. This is also true on the micro scale of personal computing, deeply considered, but we see the effects most powerfully at the macro scale of networked, interactive, personal computing enabled by the World Wide Web. The Web, freely given to the world by Tim Berners-Lee, is a metaplatform with the peculiar recursive phenomenon of unrolling before your eyes as you walk forward upon it. It is a world that appears in the very making–assuming, of course, that you are indeed a web maker and not simply a web user.

Wiener writes, “If we want to live with the machine, we must understand the machine, we must not worship the machine…. It is going to be a difficult time. if we can live through it and keep our heads, and if we are not annihilated by war itself and our other problems, there is a great chance of turning the machine to human advantage, but the machine itself has no particular favor for humanity.” If the machine is us, however, as Michael Wesch argues (and in the case of the machine of networked, interactive, personal computing on the World Wide Web, I agree), then Wiener’s statement reads like this:

If we want to live with ourselves, we must understand ourselves, we must not worship ourselves…. It is going to be a difficult time. If we can live through it and keep our heads, and if we are not annihilated by war itself and our other problems, there is a great chance of turning ourselves to human advantage, but we ourselves have no particular favor for humanity.

The idea of enlarging human capabilities should make us nervous, I suppose, but it’s a step forward to understand that that is what we’re thinking about, and that is what’s uniquely empowered and enlarged by interactive, networked, personal computing. From art to medicine to engineering to business and beyond, one capability we have and share, to an alarming and exhilarating extent, is a capability for enlarging our capabilities. Computers are an interesting manifestation of that capability, and a powerful means of using (exploiting, unleashing) that capability. As is education. (Schooling? Depends on the day and the school and the teacher.)

Once we understand that, deeply, we may to Poincare’s observation, quoted by Licklider: “The question is not, ‘What is the answer?’ The question is, ‘What is the question?'”

Licklider dreamed of using computers to help humans “through an intuitively guided trial-and-error procedure” to formulate better questions. I am hopeful that awakening our digital imaginations will lead us to formulate better questions about our species’ inquiring nature and our very quest for understanding itself.

4 thoughts on “Understanding the machine

  1. I would argue that computers don’t really learn from us. I think it’s a little dangerous to suggest they do. Only people are capable of learning, really. The people who make computers and the web learn from us. They study what we do, where we go, how we get there and make systems that facilitate our work with machines, or facilitate their companies making money off of us. If computers/machines learn from us, then you’re making learning into something far too simple.

    The machine is us in some ways. Google and Facebook especially use data from our interactions to reflect back some version of ourselves, but it is also not us, quite decidedly. It is still just circuits, programmed by humans, mostly to profit. I think we need to understand *that*. Technology is neutral. Our uses of it are not.

  2. Usually I would think that this view of the machine as ‘learning’ from us suffers from the fatal flaw of anthropomorphism. Machines exhibit behaviour that seems intelligent because they do exceedingly simple things incredibly quickly. Taken to the limit, we say they are intelligent and after a while we forget how far removed this intelligence is from our own.

    What this essay does is different. It highlights the fact that there is a symbiosis emerging between these two different ways of creating meaning in the world. The substitution for ‘us’ in place of the machine is a call to action that we need to step up our game to match the impressive strides machine data collection and interpretation. Balancing this partnership will be difficult and messy, because that is the nature of human intelligence. It will also be relentless and impassive because that is the nature of machine intelligence.

    And we thought it was difficult in the good old days, balancing relationships within our species. 🙂

  3. I agree we need to step up our game, but it’s also true that our brains cannot process data nearly as fast as computers can. So in terms of pure computational power, machines have us beat. However, I think what you mean is the idea of thinking about what purposes that data and interpretation are put to. I think the way we can collect data and “analyze” far surpasses our thinking about the sensitivity of that data and the issues having that data at our fingertips presents.

  4. Pingback: Questions and Answers | Ivy Run LLC

Leave a Reply

Your email address will not be published. Required fields are marked *