Brand and The Media Lab, 20 years after

Ah, here's the quote: Stewart Brand's The Media Lab, page 68. Media Lab's director, Nicholas Negroponte, has just described a half-gigabit channel as "effectively infinite" and asks what one would do with that much bandwidth:

The Media Lab's best and brightest stuttered. One volunteered you could have a combination of television immediacy and newspaper depth and detail available at any time through such a medium. Fine, that would occupy a teensy fraction of the bandwidth. What else? Uh, every house acquires its own broadcast capability. Uh, you could broadcast solids by having them fabricated at the receiving end. Uh, everything would be instantly available as super fax -- a daily National Geographic-quality New York Times with today's news rather than yesterday's, manufactured at the breakfast table.

That was written 20 years ago. It's tempting to take potshots at past speculation about what was then the future, but let's not. Predicting the future is hard. Myself, I'm hard put to predict the present.

On the other hand, some interesting questions do arise:

First question: What happened to the bandwidth? The Media Lab folks were stuttering 20 years ago because representatives of Bell (I mean Lucent ... I mean AT&T .... I mean Alcatel ... I mean ...) had just told them that they could run 500Mb fiber optic to everyone's home. That's around 60 megabytes/second, both ways.

There is no official definition of broadband, but current reports (such as this one from the OECD) use 256Kb as a working definition. By that measure Denmark and the Netherlands are leading the way (with Switzerland, Korea and Norway rounding out the top five). In Denmark, there is just over one broadband subscription per three inhabitants. If you factor in households with more than one inhabitant, this probably means that almost everyone who wants a broadband connection has one.

However, most of those subscribers have DSL (it doesn't say what flavor). If you're actually looking for fiber, Korea has about 9 subscriptions per 100 inhabitants, Japan about 8, Sweden about 5, Denmark about 3 and the Slovak Republic about 1. So much for fiber to every home. Most of us, it appears, are slogging along with DSL and cable. Your mileage may vary depending on a number of factors, but even the theoretical maximum listed here is nowhere close to 500Mb/s for either.

How did this (not) happen? That's an interesting story having more to do with economics than technology ...

Next question: Suppose we eventually all get a half-gigabit hookup. What could we possibly do with that? The answer seems very clear in hindsight: Blow it all on extra pixels. Late-'80s TV is one thing. HD-TV is quite another. Blu-Ray and HD-DVD play back at about 40Mb/s, easily handled by our infinite pipe, but they're not the ultimate.

Even HD-TV has visible artifacts if you look closely enough. Personally, I don't much care, but the true videophile is going to want more. How much more? Recall that IMAX is multiple gigabytes per second uncompressed, and (at a guess) around a gigabyte compressed, still over twice infinite capacity.

I doubt, though, that higher definition is why most people will want a bigger pipe. I suspect most of us who want a bigger pipe just don't like to wait. Sure, I could watch HDTV over a mere 50Mb/s connection, but who wants to wait for a program to download in real time? We want it now (because we're going to watch it later -- go figure). Now even the infinite pipe doesn't look so infinite. It's still going to take 10 minutes to download a feature film in high definition. Of course, if I can watch a movie while I'm downloading 10 more, that might take some of the sting out of it.

If you buy the "no-waiting" theory then we don't want or need infinite sustained bandwidth, but we would really like infinite peak bandwidth. Give me a gigabit pipe, and I promise I'll hardly ever use it. Or maybe I will. Usage has a way of catching up with capacity, and it seems risky, at the least, to predict otherwise.

Finally, what about the speculation in the original quote? The broadcasting solids idea has an obvious weak point: you'd need a pretty special printer to go with that (modern machine tools can reproduce solids described digitally, by any of several processes, but they're not available for home use).

As to the others: We actually do have something like the "television immediacy and newspaper depth," at least by one plausible definition, in the blogosphere and in conventional news sources. The high-def New York Times is certainly feasible. You could even print it out if you really wanted to. It's another question, and an interesting one, why no one seems to be putting out that sort of content to the mass market.

Paying for quality

Here's a music pricing scheme I've seen a few times: For one price, you get a piece of music in, say, mp3 format with a given bit rate. For a somewhat higher price, you get the same piece of music in, say, flac format with higher fidelity.

This seems a very interesting test case for the "information wants to be free" theory. Intuitively, better quality ought to be worth paying for, but on the other hand, copying either version is essentially free, so why should there be a difference?

Suppose the lo-fi version is free and the hi-fi version costs a zillion dollars. No one will buy the hi-fi version (and no one makes any money)

Suppose lo-fi is free and hi-fi costs a pretty penny. Most likely some copies will sell, but many more will circulate illegitimately. No one makes much money.

What if lo-fi costs a modest amount and hi-fi costs a bit more -- pretty much what we have in most cases? Quite possibly there will be some sort of equilibrium. Some people will pirate, but enough people will pay full price to keep the bits flowing.

My totally off-the-wall guess is that at the optimum, the proportions of pirated copies for the lo-fi and hi-fi versions will bear some simple relation to each other. On the other hand, it seems quite possible that no one will want to buy the lo-fi version no matter how cheap, if it's really bad, and conversely there's a point at which it's not worth paying for that last bit of signal quality.

The optimum price will depend on a number of factors, including local law and the cost of breaking it, demand for the particular piece of music, and (I tend to think) people's judgment as to how fair the price is and to what extent piracy hurts the record label as opposed to the artist.

When Stewart Brand said "Information wants to be free" he actually said

On the one hand information wants to be expensive, because it's so valuable. The right information in the right place just changes your life. On the other hand, information wants to be free, because the cost of getting it out is getting lower and lower all the time. So you have these two fighting against each other. [See here for example]

We don't always remember both halves of that statement, but both are essential.