Software Will Eat the Rest of the World
Alan Pentz, CEO
Washington DC
May 27, 2020 - Software’s Not Done Eating.
Nine years ago, Marc Andreessen wrote a seminal op-ed in the Wall Street Journal entitled Software is Eating the World. The piece embodied the subsequent era of cloud-based companies taking over the economy while they led a rebound in the U.S. stock markets following the Great Financial Crisis. Software enabled by rent-as-needed servers from Amazon Web Services; widespread deployment and adoption of broadband, 3G, and then 4G; and the rise of the smartphone dramatically increased the reach and profitability of software companies. One need only look at the top 10 companies as measured by market capitalization in 2008 and 2020.
In 2008, the most valuable companies were oil and gas and banks. The only software company on the list was Microsoft, milking a legacy Windows monopoly, which had yet to implement its very successful and profitable cloud strategy. By the time Andreessen wrote his op-ed in 2011, Apple climbed almost to the top of the charts and Microsoft held on by its fingernails (sans cloud strategy).
By March 31, 2020, as the COVID-19 pandemic took hold of the world, the top seven most valuable companies were primarily software companies (one could argue about Apple). Two of them were also Chinese, demonstrating the rapid rise of their tech economy. We’ve also seen many other start-up companies emerge as unicorns rocketing to billions of dollars in revenue; meanwhile, incumbent (i.e. non-software-based) players watch their businesses wither.
So, Andreessen was right? Did software eat the world? Well, not quite all of it.
In April 2020, Andreessen released a blog called It’s Time to Build, where he makes the case that COVID-19 revealed software’s failure to eat the entire world. While the price of running a web application fell 99% over the last decade, the cost of many other goods and services has done nothing but grow. Healthcare, housing, and education have continued to increase far beyond the rate of inflation and now, for many, these goods are nearly unaffordable. Additionally, our cities, transportation systems, and other infrastructure suffer from a lack of investment. Why? While the virtual world of the internet, mobility, and the cloud thrive even as it rides down the cost curve, more traditional, physical, protected, and regulated industries have taken more of a share of our collective wallets (healthcare, education) or suffered from underinvestment (infrastructure).
Andreessen’s answer for why this happened is a lack of will, a lack of desire to overcome not in backyard development or rent-seeking higher educational institutions, and a lack of desire to invest in 21st century infrastructure. I agree, but only partially. I see another chapter in an ongoing story of technological investment and progress. Software hasn’t eaten the whole world; but, it also isn’t full yet.
Boom, Bust, and Progress
Technology has a tendency to invite over-investment. If you can see the future, you want to own it. This isn’t just a modern phenomenon. Building out America’s railroads consumed enormous amounts of capital from American, British, and other investors. Many of them faced catastrophic financial losses; but, railroads became the backbone of the 19th century and early 20th century economy nonetheless. There were thousands of car companies in the early 1900s. Most of them amounted to nothing or were bought-up by GM.
As Andreessen wrote in Software is Eating the World, many of the businesses thriving in 2011 had predecessors who failed spectacularly in the early 2000s Tech Bubble. Pets.com failed but Chewy.com is a nearly $18B market cap publicly traded company today. The world didn’t have the broadband, mobility, and logistics networks or the penetration of those networks it needed to make Pets.com a success in 2000; but, it did by 2019.
I think a similar dynamic is playing out today. The most successful companies over the last decade have structured their operations in a virtual context and have reaped the benefits: Netflix, Google, Facebook, Amazon, etc. The unicorns that emerged from the deluge of cheap venture capital during that time are a more mixed story. In general, the more virtual the business model, the more profitable the company: Anaplan, Roku, Smartsheet, Twilio, etc. The unicorns that lagged tended to mix software with higher cost elements of the physical world. Uber and Grubhub are the canonical examples of this phenomenon (not to mention, WeWork, which might not survive this downturn)
The Baumol Effect
Making profits in the physical world is harder and these are exactly the places Andreessen wants us to invest. This is particularly true when the physical world involves labor. Paying all those Uber and Grubhub drivers is expensive, not to mention, the people who run the restaurants. Much of the inflation in education expenses is driven by the proliferation of administrative staff and the lack of productivity growth in other staff. The story in healthcare is more complicated. But, the lack of growth in the productivity of labor is at least one of the primary drivers in increasing costs. People, in contrast to servers, don’t like it when you run them down the cost curve.
One reason for this is something called the Baumol Effect. When a company can make something cheaper and/or more efficiently, it can lower the price it charges and pay workers more at the same time. That’s the beauty of productivity growth. But, that also creates wage and cost pressures in industries that aren’t experiencing productivity growth. As the economist William Baumol observed, it took the same number of musicians the same amount of time to produce the same string quartet when Beethoven lived as it does today. But, today, the musicians get paid a lot more and we all pay a lot more for the tickets. Why is that? Because many other things became cheaper and more productive.
For example, as computers get better and/or cheaper every year, some of that excess wealth goes to wages for computer executives. Pretty soon, the string quartet has to pay more too or else all the musicians will become computer executives. Workers in less efficient sectors will leave those sectors if their wages don’t rise as well. But, that is only part of the story. The other factor that increases the Baumol Effect is that you can only buy so many computers before you have two in every room of your house and the last thing you ever want to see is another computer. Those computer executives start thinking about how great it would be to put their kids into fancy schools. Johnny can’t go to community college after all. He needs to express his creativity at NYU Film School and NYU only takes so many people because it isn’t getting more productive; so, the costs go up.
That’s the problem with the sectors where we need to build; they suffer from a cost disease. They aren’t becoming more efficient, and as the rest of the economy becomes more efficient, the more wealth there is to bid up the costs of the desirable but non-efficient services.
I don’t dismiss the regulatory factors. While they contribute, I think they could be overcome if productivity exposed them as the only cause for increasing costs. “You can’t build in my backyard,” pushes up housing costs but construction’s lack of productivity is a major contributor to the issue as well.
The Solution: 5G
I think the boom we saw in startups that brought more technology and software to the physical world in the 2010s was the harbinger of the future (just as Pets.com was the harbinger of Chewy.com). 4G enabled Netflix, Spotify, Zoom, and Instagram. It also enabled Uber and Grubhub; but, they just couldn’t turn a profit. 5G will solve many of these problems. (Note: I am using 5G as a catch all term for many technologies associated with the 4th Industrial Revolution: AI, Virtual Reality, 5G, Machine Learning, etc.) If we succeed in developing autonomous vehicles, Uber will suddenly be profitable because it won’t have to pay drivers. Similarly, if we build out a network of ghost kitchens, Grubhub can cut out the restaurant and the driver.
This also applies to your doctor. If a surgeon can operate remotely through robotic machines on a patient hundreds or thousands of miles away, their productivity goes through the roof. Additionally, if that doctor can replace 40% of their job using an AI bot to answer basic questions or diagnose ailments, etc. then, the doctor can spend 40% of their time on higher-value projects. That’s how you increase productivity.
If a college professor were to stream their course to thousands or even millions of students, their relative productivity would skyrocket. While a Coursera course today might seem like a pale reflection of your seminar class on 19th century French poetry at Reed College, it’s only because the technology and tools are 4G. 5G will create virtual experiences, unlike anything we observe today. So, we might not replace Reed College but couldn’t we make Penn State open to many more students at a vastly reduced cost?
Personally speaking, I see another area where 5G could improve my life. My wife doesn't like getting online groceries delivered and I have to agree that it just doesn’t work that well. The store is inevitably out of things. The shopper substitutes the wrong kind of apples. If I’d been there I might have gotten pears instead of another kind of apple. Add to that the fact that I still have to be online, approving or not approving changes, and it makes for a pretty terrible experience. I’ve had similar glitchy experiences with restaurant delivery. It just isn’t the seamless experience you want.
But what if the grocery store only did delivery? It’s inventory system would be built into the app so you never selected something the store was out of? And what if your groceries were gathered and packed automatically by a robot that used machine learning to learn that you prefer Bosc pears to Macintosh apples if they are out of Fuji? What if the store messaged you when you needed things because it tracked your consumption habits and what if that delivery were cheaper than actually going to the store? I’m guessing we’d switch.
The pandemic forced us to try online delivery. For every old school family like mine, there’s another that is learning they could live with the glitches. That brings in new customers and revenue but it’s also spurring massive investments to improve. I’m guessing we just accelerated the grocery delivery service by a decade. 5G will allow software to eat much more of the physical world. Industries that have been safe until now: healthcare, education, construction, etc. will be disrupted. This will be great for some like the college professor who will make three times as much serving 100 times as many students. It’ll be not so great for others like the other college professor that no one needs anymore. Most importantly, it’ll be infinitely better for society as a whole by bringing these high costs services to more people.
So, I agree with Andreeson that we need to build and what we need to build is 5G and the technologies that it will enable. It’s time to enter the next phase of technological development. It’s also crucial that our government partner with the private sector by investing in that 5G infrastructure and R&D to ensure that the U.S. leads the world in these new industries. That kind of investment will deliver the results implied by companies like Uber and WeWork and it will deliver a new American century. It’s all a part of the inexorable advance of technology. Software just had it’s mid-morning snack and lunch is the biggest meal of the day.
Author
Alan Pentz, CEO and Founder of Corner Alliance, has worked with government leaders in the R&D and innovation communities across DHS, Commerce, NIH, state and local government, and the non-profit sector among others. He has worked in the consulting industry for over ten years with Corner Alliance, SRA, Touchstone Consulting, and Witt O'Brien's. Before consulting, Alan served as a speechwriter and press secretary for former U.S. Senator Max Baucus and as a legislative assistant for former U.S. Representative Paul Kanjorski. He holds an MBA from the University of Texas at Austin.
