At the end of December 2009, I published a set of predictions about both my personal life and what the technology industry would look like a decade later—in 2020. I got a lot correct, I’m proud to say, especially about the big picture of social issues resulting from technological change.
But I also made some big mistakes in forecasting, from which I hope to learn for the next decade.
On the plus side, I called correctly that by 2020 Facebook and Amazon would be big antitrust targets. I also correctly identified that privacy would be the major societal issue with technology. I would argue that as I predicted, we are starting to talk about the authoritarian implications of centralized web services.
My predictions about the hardware and service stack of today generally turned out to be correct. As I predicted, I still primarily use a MacBook and Google services. I got the deep integration between IM and SMS, which clearly is dominant in the form of iMessage. I correctly called the shifts toward cloud and subscription for media and predicted that the best content creators would be in a golden age.
I even called that Venmo (which was a year old in 2009) would be the dominant social payment product in the U.S. I got right that content monetization would be all personal targeted advertising: real-time generated personalized bids and offers.
There are three big thematic things that I clearly turned out to be very wrong about.
I was too optimistic about the openness and commoditization of certain parts of the technical stack, including search and messaging. Some of the technology behind these services has indeed been commoditized, but not the services themselves.
And while I got the huge impact of passive big data and signals, I thought there would be far more consumer utility stemming from that future rather than back-end uses for targeting and algorithmic optimization.
Finally, I was far too pessimistic about the U.S. economy in the wake of 2008. I turned out to be dead wrong in my predictions that in 2020 we would have very high unemployment and extremely expensive energy.
The call I am most embarrassed to have made in 2010 was that I would be taking 200 photos per year, when in reality I probably take 200 photos a week. My rationale for 200 photos a year was an expectation that there would be a strong shift toward video. I massively underestimated the degree to which recording memories in all forms would swamp any mix shift toward video and away from static photos.
A good lesson from history, which I now know, is that legacy technologies tend to continue to grow for years even after the emergence of the technology that will ultimately supplant them. For example, the horse population in New York continued to grow long after the introduction of the automobile.
Lessons Learned
Looking forward and considering the predictions I want to make about 2030, the following are the specific lessons I will be trying to keep in mind:
(1) Consumer hardware and software platforms have very high systematic lock-in and evolve much slower than people generally realize.
I correctly called in 2010 that my primary experience of technology wasn’t going to change much in the decade. To be sure—as was easy to predict—everything got better and more powerful, but fundamentally my laptop is still a MacBook, my phone is still an iPhone, and Chrome is my default experience of the web. I spend most of my working time in Google products and most of my social time online on Facebook. The layers of consumer hardware and software were already locked in back in 2010. They are only more so today.
(2) Primary messaging behaviors and services also haven’t changed much and I wouldn’t expect them to evolve much going forward. They are stickier than most people realize, and the messaging disruption of the early 2000s was mostly a result of mismanagement rather than being the natural course of things.
In 2010 I correctly predicted that email would continue to be my primary mode of communication in 2020 (it is and will remain so). I am still using the same messaging services more or less that I was in 2010 as well (though I get that it’s not the global norm). I also got that videoconferencing was clearly the big change of the next 10 years—and that I would be consuming much more of it—and I am. I believe that this trend will accelerate and that virtual reality is today nearly where videoconferencing was in 2010.
(3) In my predictions I got a lot right about the dominant services of 2020, but I thought that 2010’s dominant services (like Google, messaging interfaces, and even AT&T) would get commoditized and become “open” within the decade. This turned out to be wrong. One can argue that many other players have caught up to the the quality of Google’s core search results over the last ten years, but well-run organizations are able to keep lock-in going regardless.
The commoditization of a technology does not necessarily lead to the commoditization of the company that dominates it.
(4) My attempts to give credit to things I saw as rapidly changing didn’t go far enough. When things are moving, they move further and faster than you might expect.
For instance, I saw that from 2010 to 2020 there would be a huge move of personal data into the cloud. But I expected that my personal data would move from 30% in the cloud to 70% in the cloud. It turned out that 99% of my data is now cloud hosted.
This is also where I underappreciated how far the rise of videoconferencing would go. I expected to do five times more video chat by now, but it is probably more like 20 times. It is also where I massively underestimated the amount of photos, videos and so on I would record yearly. I clearly understood the move toward video, but not the rate at which all recording would rise.
(5) At the same time, I was too optimistic about “edge” technology. I expected that by 2020 robotics, nano, bio and health technology would start to be meaningfully deployed. While you can squint and argue that elements of these technologies are in play in our everyday lives, the wonders they theoretically represent are not yet present.
The lesson is that if a technology doesn’t yet exist, don’t bet on its realistic use within 10 years.
(6) While technology might be less cyclical than expected, don’t forget the larger set of forces at work in the macroeconomy.
The place that I went too far on expecting disruption was the economy. Coming off of 2009, I was very pessimistic about unemployment in 2020 because of technology, and I expected very high energy prices. These are places where I think I wasn’t being sensitive enough to the cyclical nature of our system and the power of governments and regulation. In technology I expected more “cycle” characteristics between open and closed systems than played out. But in macroeconomics I made exactly the opposite mistake—I expected the trend then underway to continue. But I was wrong.
(7) On a related note, beware recency bias. It is clear to me that many of my macro bets were on some level tied emotionally to the 2008 financial crisis. This bias is well known but nonetheless it is critical to look out for it.
(8) On a personal note, don’t underestimate the impact children will have on your life. I didn’t fully internalize that my children would become my No. 1 priority—over myself. I take far fewer flights, and I don’t do time-consuming adventure sports (biking, skiing) as much as I thought I would because of them.
Interestingly, I also didn’t realize how positive it would be for me to do strength training. So I am healthier than I thought I would be even with children. This isn’t a mistake I will make again (but let this be a possible lesson to people currently in their 20s!).
It is easy to lie to yourself about your past beliefs and perspectives. I find the discipline of publicly committing to opinions, and then looking back and holding yourself to account on what you did and didn’t think, one of the most useful things to do in evolving my own perspective.
In total, when I look back at my 2010 set of decade predictions, I can’t help but feel they are quaint. I was trying very hard at the time to think big. While I got some of the biggest themes right, so many of my predictions now seem small and irrelevant. That is a testament to how much the technology world has come to dominate our whole lives and the global economy in the last 10 years.
Technology was an “industry” a decade ago, whereas now it is everything. The stakes have never been higher and likely the predictions have never been harder than they will be to make in the next decade.
If you are interested in how these lessons inform my next set of predictions, I would invite you to review (and debate if you like) my 50 predictions for 2030.