Skip to content

Inscrutably Obvious: A Paradox of Technological Change

7 minutes read

A case study from the cable industry at the dawn of the internet.

In 1997, everyone thought the cable industry was dead.

The cable companies had invested extraordinary sums building high-speed electrical piping into homes around the country. The cable industry had been promising "interactive TV" for years, but it wasn't really coming. Cable had already been installed in more than 90% of homes, but the additional infrastructure required for interactive TV turned out more difficult and expensive than expected. Cable stocks were at an all-time low, and an unpredictable regulatory context made matters even worse.

Today, with the benefit of hindsight, it's obvious that cable was going to be the perfect physical transmission belt for the internet. But surprisingly, as late as 1997, the idea was hardly considered.

You see, Wall Street had given up on the cable companies. Wall Street feared that cable wouldn't survive at all, so they could hardly fathom that cable could be poised to support an impending technological revolution.

The first person who understood this was Bill Gates, who invested $1 billion to buy 11% of Comcast in 1997. But this was an idiosyncratic and contrarian bet, inspired by a chance encounter.

When the idea first came up—at a dinner party with cable executives visiting Gates in Silicon Valley—it came up as a joke. A few days later, Gates called Comcast in earnest to make the deal.

🗄️
The facts and figures from this post come from the book Cable Cowboy: John Malone and the Rise of the Modern Cable Business by Mark Robichaux.

So why was cable not widely perceived as the next step for the internet, until Gates bet on Comcast in 1997?

I'd like to suggest that sometimes massive and imminent technological changes are inscrutable precisely because they are obvious.

Paradoxically, we often find it more comfortable to believe outlandish theories of the near-future than highly logical theories rooted in clear observation of multiple converging trends. Outlandish ideas about the future flatter our pride. We like to think we are creative, that we can see things that don't yet make sense in the normal reality of average people. Outlandish theories have another psychological attraction: Disconnected from reality, they're usually not actionable or investable, which means professing them is a cheap way to show off one's creativity.

Strong theories of technological change are painful to hold. They're not impressive at cocktail parties; in fact, they are more boring than sports and weather. Though the implications may be remarkable, the theory often reduces to a collection of uncontroversial statements about uninteresting things (and to get the implication you have to work through all the boring premises).

Well-grounded theories are also terrifying to hold because their groundedness means you can usually find a way to bet on them. If you can bet on them, that means you can turn out wrong, and face negative consequences, which is stressful.

For all of these reasons, it is possible we are most averse to the strongest theories about the biggest changes. Such theories will be the most excruciating to express when they're still early, and at the same time, they will require of their proponents the most frightening gambles. If the theory is really so strong, and the change really so big, then shouldn't you bet everything on it? If you don't, you must be a dummy or a coward.

Thus, when we encounter a potentially strong theory about a big change, most people would rather look away than allow themselves to become convinced.

This is how the obvious becomes inscrutable.

When Computers Sang

For the average internet user in 1997, download speeds were terrible. That's because the internet operated over phone lines, which are much thinner and slower than cable lines.

My Millennial readers (and older) will remember the distinctive sound of America online "dialing-up."

audio-thumbnail
AOL Sign On Dial Up
0:00
/0:30

Back in 1997, America Online was jokingly called America Waiting Online. The website prefix www was thought to stand for World Wide Wait.

While the narrative of "interactive TV" was considered a total failure and dismissed as hype, Gates saw that the vision was not totally wrong. It was just incorrectly framed, and a little too early.

Gates saw that households would indeed be interacting with rich data on screens, but those interactions would be facilitated by internet service companies rather than television broadcasters. Gates wanted Microsoft to provide the end-to-end software system for America's cable system.

What's most remarkable is that—at this time when cable seemed down bad—from a slightly different angle, cable was invincible. The cable companies owned the fattest and highest-bandwidth pipes connected to American homes. All of the new internet companies really had no other chance of getting into American homes at any kind of scale.

If we can better understand the phenomenon of inscrutable obviousness, perhaps we can illuminate opportunities hiding in plain sight today.

Jaded by Jumping the Gun

It's well known that we tend to interpret new technologies in light of the old (cue your favorite hackneyed example, like how many people thought the early PCs were just fancier typewriters). But there's a slightly more specific variant of this tendency, which is less discussed.

We irrationally over-index on the earliest narratives that emerge around major technological changes.

If the first big narrative proves wrong, we find it very difficult to reject the falsified narrative while also believing, with equal confidence, a totally new narrative. If the first big narrative fails, our bias is to infer that because we were wrong in the specifics, we must have been wrong in general: "Cable was all a bunch of hype, forget about it!"

What we now call cable was originally a way for rural cowboys to bring the big coastal TV networks to remote areas via antennae. If you set up a big antenna, you could catch the TV airwaves, and then run cables from the antenna into the homes. These early cowboys made decent fortunes running this kind of community operation.

By the time of the "interactive TV" narrative, cable had already gone through one massive narrative change. When pioneering content companies realized they could build subscription businesses by shipping exclusive content through cable pipes, this play was no more of a consensus than Gates' investment in Comcast years later. The pioneers of this first narrative revolution were companies like HBO and CNN, massive winners in the golden age of cable TV.

By the time the interactive TV narrative flopped, it must have been really difficult to believe that this massive national skeleton of cable piping could be used, yet again, for something completely different and unforeseen.

"When I turn on my modem, the computer sings to me.” —Björk in 1995

At any given moment in history, inferring the next stage of a general technology requires patient study and a scrupulous accounting for anchoring and recency biases. It also requires concrete analysis of specific players and their medium-term incentives and costs. Gates was able to see that cable was the future of the internet, but only from his own independent analysis and abstract reasoning about the specifics of the situation. There were many other players in tech with big war chests who could not determine what Gates determined.

Too Good to Be True

I would proffer another thesis about the determinants of inscrutable obviousness. Though this thesis is not necessarily evidenced in Robichaux's book, it's something I've written about previously.

When we benefit from a particularly remarkable and rapid technological innovation, we tend not to see how much better that innovation could still become.

A service like AOL was so magical compared to the pre-AOL world, that lackluster download speeds were probably naturalized. Maybe people joked about the World Wide Wait, but they joked in a complacent resignation. When dial-up internet was slow, we inferred, "OK, I guess this new thing is just a slow thing." At that time, there was no evidence that people needed it to be faster because it was already so much faster than... not having the internet at all.

In 1997, if you said it would be better to run the internet through the cable system, in order to make the internet faster, most investors would have said: "What makes you think people want or need the internet to be faster? People are hardly maxing out the known use-cases; there's no evidence people want to do things like run high-bandwidth work meetings over video calls!" In other words, dial-up was slow, but in the context of a revolutionary technological innovation, it's hard to see that an irrepressible force would require it to get faster. It's hard to see the telos of a new, general technology.  

I find the lesson of cable highly salient today. All of our "personal computers" are owned and neutered by large corporations and server mafias, but we've naturalized it. It seems ridiculous to propose that computerized social networking could or should be totally redone in a better way. Yet the lesson of cable is that betting on paradigm change—if the logic is tight enough—is sometimes the smartest bet in the world, even when it makes no sense to many educated observers.

We're seeing the same issue with crypto. We now have fully automated and collectively owned software protocols worth billions of dollars, yet during the present market slowdown, I'm seeing countless educated observers proclaim the death of crypto. "See? It was all a bunch of hype," they say.

It seems crazy to propose that we could cross-apply other existing technologies and make something significantly more powerful than the status quo of personal computing, or crypto, or both.

But if you sit down for patient study and you account for all of your biases, what do you think is the next stage of the internet, personal computing, and crypto? What can we see right now, which is invisible to most people right now, but will be seen as inevitable 10 years from now?

Next

Subscribe to receive the latest posts in your inbox.