OVER THE EDGE
I like Edge.org because I disagree with almost everything they publish. This gives me a warm, satisfied feeling of intellectual superiority without having to do any difficult work.
Edge participants are generally very enthusiastic about scientific culture. This is not a bad thing, of couse. But sometimes their enthusiasm causes their speculation about the future of technology to outpace even the goofiest utopian sci-fi pulp.
Case in point: The sunny optimism about science’s potential for improving our lives, without any real understanding of people’s actual lives, makes this Danny Hillis’ article sound like something out of that 1950’s Popular Science “Life in the Year 2000!” article.
I can’t find a scan online, but it famously predicted that the housewife of the future would clean the house by spraying it down with a garden hose, because everything would be waterproof. Not only did this not come about, but the naivete about which parts of daily life are maleable and which are not is charmingly quaint in retrospect. In fact, culture has been by far the largest area of change in the last 50 years. A dynamic and changing culture is the medium in which technologies either flourish or perish.
The inscrutability of social factors when making predictions for the future is not a problem exclusive to scientists and engineers, but Hillis provides a pretty amusing example with his ideal tutor automaton, Aristotle.
“First, imagine that this tutor program can get to know you over a long period of time. Like a good teacher, it knows what you already understand and what you are ready to learn. It also knows what types of explanations are most meaningful to you. It knows your learning style: whether you prefer pictures or stories, examples or abstractions.”
Ok I’m imagining it… no wait, I was thinking about pie. Ok I think I’ve got it now.
“Imagine that this tutor has access to a database containing all the world’s knowledge.”
No. I cannot imagine an actual database containing all the world’s knowledge. Who is going to type it all in, along with it’s various stories and illustrations? How many people will it take to maintain it? How long will a search take? Oh right, I won’t need to search because this database will know what I’m ready to learn.
“I will call this database the knowledge web, to distinguish it from the database of linked documents that is the World Wide Web.”
I will call this database the impossible web, to distinguish it from the actual web and anything that could ever possibly exist anywhere.
“Given such a database, it is well within the range of current technology to write a program that acts as a tutor by selecting and presenting the appropriate explanations from the database.”
There are two things wrong with that sentence. I suppose if God came down and handed us a database containing a complete list of facts, it wouldn’t be impossible to extract information from it in a pedagogically useful form. The other problem is that it is not remotely forseeable that a tutor of the sort we were asked to imagine. Yes, computers can learn your preferences by crunching the statistics on past decisions and behavior, (via neural networks or Bayesian algorithms, for example). This is a far cry from providing the kind of personalized instruction Hillis imagines.
Next, Hillis takes us on a magical mystery tour of what problem the knowledge web might solve.
For example, imagine yourself in the position of an engineer who is designing a critical component and wants to learn something about fault-tolerant design. This is a fairly specialized topic, and most engineers are not familiar with it; a standard engineering education treats the topic superficially, if at all. Fault-tolerant design is an area normally left to specialists. Unless you happen to have taken a specialized course, you are faced with a few unsatisfactory alternatives. You can call in a specialist as a consultant, but if you don’t know much about the field it’s difficult to know what kind of specialist you need, or if the time and expense are worth the trouble. You could try reading a textbook on fault-tolerant design, but such a text would probably assume a knowledge you may have forgotten or may never have known. Besides, a textbook is likely to be out of date, so you will also have to find the relevant journals to read about recent developments. If you find them, they will almost certainly be written for specialists and will be difficult for you to read and understand. Given these unsatisfactory choices, you will probably just give up.
The first option seemed pretty good to me, actually: call a specialist. What’s the problem with that? If you’re not a specialist, you don’t know what sort of specialist to call! Catch-22! (Also: the system won’t be out-of-date like a textbook, yet will not require the professional knowledge and linguistic competence assumed by research publications. I guess even internal consistency is too much to ask.)
The normal human reaction to this sort of scenario is to ask the opinions of those who have had experience with the relevant issues. People love to give advice, and the currently-existing World-Wide Web (or Usenet for that matter) is built on the sort of free advice-giving exactly suited to resolving the dilemma that Hillis poses. Unfortunately, the actual web contains bad advice. The solution being suggested is that we should build a web that only gives good advice.
So let’s explore further what the web of good advice can do for us:
- “Aristotle would show you a map of what you need to learn … by comparing what you know to what needs to be known to design fault-tolerant modules.”
- “Aristotle knows what you know because it has worked with you for a long time.” (But what about before it’s worked with you for a long time?)
- “Aristotle might take your word for what you know, but it is more likely to quiz you about some of the key concepts, just to make sure.” (Well then Aristotle can go fuck himself.)
And that’s only in the first of several paragraphs of breathless flights of fancy about Aristotle’s amazing possibilities.
Every single paragraph contains claims more absurd than the last. I desperately wanted to find some sort of recognition that Aristotle is really just a fantastical thought-experiment, but I instead found the opposite: “Such a Primer [an AI from a Neal Stephenson novel] is beyond the capabilities of current technology, but even a program as limited as Aristotle would be a step in that direction.” Holy Christ. I can’t continue. It’s all just too stupid.
Artificial Intelligence is dead dead dead if these are the kind of ideas seriously being suggested. You’d think 50 years of abject failure would have engendered some humility.