Can Tech Die

I can’t recall when we all first began to collectively refer to the computing-powered high-technology sector, based primarily on the US West Coast, as simply “Tech” (I’ll drop the scare quotes but keep the capitalization for the rest of this essay). I think it was shortly after the 2000 dotcom crash. I do, however, have a theory about why we started doing that, and why we might soon have to stop.


Tech is obviously not the only locus where technology work, or even high-technology work, gets done.

Even the idea that Tech dominates software is shaky. I recall someone pointing out (this was admittedly about 15 years ago) that Lockheed Martin, an old economy aerospace company, writes and manages more code (and more complex code) than almost all Tech companies. But nobody has Lockheed Martin in mind when they use the word Tech without qualification.

On the other hand, Tesla is Tech but Ford is not. The former fundamentally views itself as a computer company that builds cars, while the latter views itself as a car company that might use computing (including advanced computing for driverless/EV car projects).

Other weirdness: IBM’s membership in Tech is increasingly suspect, while many companies that seem to be doing non-computing-centric things like synthetic biology and drones seem to qualify.

Apparently, a prominent role in the history of computing is no guarantee of inclusion in Tech. And not being primarily about computing or software is not a certain disqualification.

So what’s going on here?

What Makes Something part of Tech?

Though it’s tempting to conclude that the use of the term Tech in such a narrow way is arbitrary parochialism on the part of Silicon Valley (plus major outposts here in Seattle and a few other places), that’s not the reason.

Not least because everybody, not just the people within Tech, participates in the consensus to call it Tech, seems okay with the term, and has fairly good pattern recognition around what belongs in the set and what does not. Nor do technologists outside of Tech seem to particularly mind the apparent appropriation.

The charge of parochialism is also simply not true at least within technology in a broader sense. Most good software technologists I know are also generally interested in all kinds of technology and engineering, going on anywhere in the world. Technologists outside of Tech are also generally interested in the technology of Tech, and in learning from it.

What matters in whether or not something is part of Tech sector is not how much, or what sort of technology work is going on in a sector, but who drives it.

The dominant feature of Tech is that technologists, rather than sales and marketing people, or politicians, are generally in the driver’s seat.

Even when sales and marketing people are apparently in the driver’s seat, what’s driving them is demands from technology leaders in customer companies. One way or the other, every important decision in Tech is ultimately made by actual technologists (whether well or poorly is another matter). Even in the much-maligned world of AdTech, the algorithmic foundations are complex enough that engineers, rather than advertising professionals, end up making most of the key decisions.

All Tech is One

In applying this theory of Tech, you have to ignore organizational and product boundaries, and look at entire ecosystems and stacks, along lines of interoperability, OEM relationships, talent mobility, and M&A activity.

Tech is an overall pattern of strongly internally entangled economic activity by means of which pieces are dynamically bundled together to create services and products actually used by people. End-user capabilities emerge of the growing soup of deeply interconnected potentialities that is the Internet.

For instance: buying something might involve all 4 of the big Tech companies (see a recommendation on Facebook, search for reviews/blogs on Google, buy it on Amazon, from an Apple or Microsoft device).

Theories like Ben Thompson’s Aggregation Theory help track the details of how and why this plays out, but the headline is: Tech is a single, connected, hydra, within which technologists make almost all the decisions that matter. Because they’re by definition the only ones keeping up sufficiently with the nature of the potential coming online, and thinking about what to do with it, to make decisions at all.

And if you learn enough to meaningfully participate, hey, you’re a technologist too. You’ve been assimilated by the Tech Borg.

The existence of a specialized VC sector is a sign that an area of technology is part of Tech. When technologists are in the driver’s seat, patterns of return on capital acquire a certain generic predictability (in terms of time horizons and rates) that allows for an efficient kind of investing.

A good sign that a technology is not part of Tech is that it does not fit the venture capital investment profile. This generally means it is evolving too slowly, or with too low a return rate, for technologists to corner most of the decision-making authority. That in turn means other actors will muddy the picture, creating unmanaged uncertainties in risk profiles, time horizons, and return rates, making VC-style investment harder.

When you wander out of Tech proper, you better have clever alternative investment models and vehicles to work with, because the VC playbook won’t work (something that a lot of idealistic technologists who want to solve “real problems” find out the hard way, after they fail to raise VC money for their — often very good — idea).

When the economy is in a Tech phase, it almost doesn’t matter if non-technologist actors in the economy — lay people, labor leaders, politicians, courts, religious leaders, entertainment celebrities, environmentalists, humanists — act successfully to regulate bits and pieces. For every attribute of Tech where technologists cede some control, two more attributes seem to come online that non-technologists don’t even know about until it’s already too late.

For every domain fenced off from encroachment by tech, three others are willing to let it in to gain an advantage over the fenced-off domain.

Tech is a growing agency pie, and a Techie could be defined as someone who is participating in a way that they’re gaining agency faster than secondary actors can take it away. So Tech is simply the collection of all such Techies in a densely connected social graph participating in making the first decisions about what to do with emerging new technological potential.

It is when this process slows, and others begin to catch up and infiltrate the Techie network, that Tech becomes Not Tech.

Tech is not technological activity per se, it is a rate regime of technological evolution, defined as “fast enough”, and a necessary and sufficient condition to be a technologist in Tech, or a Techie, is simply being able to keep up.

How fast is fast enough?

When 95% of the population is being left behind, complaining “too fast!” The bulk of the population being left behind is not a bug in Tech as a social process, it is the defining feature.

When they stop complaining and start acting en masse, that’s when it’s no longer Tech.

What Makes Something “Not Tech”?

Companies that are not part of Tech, no matter how much technology they build, and how advanced or software-driven it is, have all consequential decisions made by non-technologists. This is because new agency is not being created by technologists faster than it is being taken away from them.

For example, in the defense sector, politicians decide what weapons systems to buy, and fight over where to build them. The weapons systems are merely politicians’ means to other ends, like creating jobs. Even military professionals, let alone technologists, don’t get as much say in the key decisions as they’d like.

In the US healthcare industry, decisions are made by politicians, insurance companies and doctors, in that order.

In education, decisions are made by teachers (often unionized), politicians, and parents, in that order.

In Hollywood, decisions are made by financiers, Chinese leaders, and creative artists of various sorts. In that order.

In general, the more mature a sector, the more key decision-makers will look like either financiers or politicians. Agency over technology flows from technologists in Tech to financiers and politicians via a sort of (American) football-shaped curve, where the width of the football determines the diversity/range of decision-makers (who bring with them a diversity of interests and expectations across a growing range of time horizons and patterns of “return”).

We’re starting to enter the middle of the football now, with Tech losing ground to Everybody Else. Some pieces of Tech have skipped the middle entirely and landed straight in the endgame, with bankers and politicians vying for agency.


Software Eating the World

Why do we say “software is eating the world”? Arguably, “healthcare is eating the world” is a more justified statement, since it is gobbling up a growing share of the economic pie.

Or perhaps “finance is eating the world”, since more and more of the developed world economy in particular is starting to look like finance by other means. Every kind of value is being financialized, every kind of asset is securitized, and every large company looks more and more like a bank with every passing year.

The thing is, while other kinds of decision-makers occasionally enjoy phases of monopolistic decision-making agency over large swathes of the economy, they have no real active means to drive the rate of the process by which they gain agency. They can’t invent a bigger pie for themselves. The best they can do is resort to fraud for a while.

They have to wait for Tech to mature (ie slow down enough) and the game to come to them. Healthcare eats everything in proportion to increasing longevity and aging populations (or fraudulent medicalization of naturally healthy conditions). Finance eats everything as other kinds of risks are squeezed out and the sector becomes ripe for “harvesting” through financialization (or of course, fraudulent “growth” again).

During the installation phases (cf: Carlota Perez) of major world-eating technologies driven by Moore’s Law type exponential dynamics, Tech has a unique potential for proactively growing its power over world history. Software eats the world at the rate engineers figure out ways to apply the potential of the underlying fundamental technological improvement curve (Moore’s Law).

Something similar happened with printing, with steam, with electricity, and with oil. And now it’s been happening for 20 years with software.

In the last 400 or so years, there’s always been a relatively active “eating” tech at work in the world, being driven by the technological imagination, limited only by how clever people are capable of being, and driving the history of the world faster than any other competing force.

In one sense, Tech is just a faster-than-human phase in the evolution of new technological capabilities, when the internal constraints and capabilities of emerging technologies serve as stronger forcing functions and levers of value creation than any competing external force.

Or more precisely, in a Tech phase of technology, the impact of the right engineering decisions overwhelms the impact of almost all other decisions — social, financial, marketing.

Faster processors, lower-latency networks, more memory, lower bug rate, faster bug resolution rate, declining cost curves (Moore’s Law being the big one), automation feeding on itself: all these features of technology in a Tech phase conspire to empower engineering decision-making over other kinds.

If you get every other kind of decision wrong, but get the engineering decisions mostly right, you can still win. On the other hand, if you get the engineering decisions wrong, but every other kind of decision right, you’ll likely lose.

What’s more, the success or failure of individual products or companies is almost irrelevant. It’s all one giant soup of improving engineering, and anything valuable developed anywhere has a decent probability of migrating through the ecosystem and finding a good home in some product, in some company, or in the open-source world. Give or take a few years and a few more enabling conditions to click into place somewhere in the overall ecosystem.

Technology is Tech during the periods that it is single, connected, indefinitely sustainable game driven by an open culture of strong engineering decision-making overwhelming everything else.

This is neither good, nor bad. It just is, and soon it might not be anymore.

Can Tech Die?

Tech does not every really die, except with civilizational collapse and species extinction. It can, however, go dormant for long periods of time, especially within specific geographies.

But for the last 400 years or so, Tech has never entirely gone dormant in the world at large. When China quit the game, Europe took the lead. When European leadership flagged, America took over. Some think that now that American leadership is starting to show signs of fatigue, China will take over the lead again.

I have my doubts. One big reason is that while China (the state) can drive technology, it cannot allow Tech qua Tech to exist as a Promethean, ungoverned force in the world. Ironically, even though Chinese political leadership is dominated by engineers, they are the opposite of technologists. They are technocrats. Technology-powered bureaucrats who want to slow technological evolution to a presumed-natural human pace. Chinese leaders may want software to eat the world politely with chopsticks rather than forks, but that is a “with Chinese characteristics” distinction without a difference as far as Tech is concerned. Software going to finishing school to learn table manners for eating the world.

So China won’t save Tech. The rise of China is in fact a sign of the slowing of Tech.

Tech can’t die short of human extinction, but it can go dormant everywhere for a time — a long time — bringing on a sort of technological Dark Age ruled by non-technologists. You don’t need technological evolution to stop for this to happen. It just needs to slow down enough that non-technologists can take away agency faster than technologists can create it.

If and when then this happens, “breaking smart” will no longer be an accurate descriptive phrase for participation in the leading edge of technology, because it won’t be evolving fast enough to be Tech. When that happens software won’t stop eating the world, but it will go from wolfing it down in huge gulps to nibbling politely, using the right forks or chopsticks for everything.

The process may be smart, but Tech won’t be breaking smart from its own past and the history of the world, creating a gap between the necessary and the possible for the imagination to flourish and serendipitous amounts of new wealth to emerge.

Some would see that as Golden Age, one where finally things have slowed down enough that devil-possessed technologists have been put in their place by wise artists, environmentalists, lawyers, politicians, and bankers.

I think it would be a tragedy of course. One that has already begun to unfold in many parts of the world. Because when technology stops eating the world, people start eating each other.