Machines Taking Jobs: Why This Time Is Different

Will AI and robotics revolutionize human labor or not? 

More than half of all US jobs could be disrupted by automation in the next several decades; at least that’s our opinion. About half the people we talk to disagree. Those that disagree think AI will open up new job opportunities by enhancing human abilities. A common element to their argument is that we’ve always had technical innovation and human work has evolved with it. A few examples would be the cotton gin, the printing press, and the automobile. All of these inventions threatened jobs of their era, but they ended up creating more jobs than they destroyed. So why is this time different?

Because, for the first time in history, we don’t need to rely on human intelligence to operate the machines of the future. The common denominator among those three examples and countless other technical innovations is that they were simply dumb tools. There was no on-board intelligence. Humans using those tools provided the intelligence layer. Humans were the brains of the cotton gins, printing presses, and automobiles. If the human operator saw or heard a problem, they fixed it and continued working. Today, the intelligence layer can be provided by computers through computer vision, natural language processing, machine learning, etc. Human intelligence is no longer required.

You might say that machines aren’t nearly as smart as humans, so they aren’t as capable as humans. But in reality, they don’t need to be. AI required to operate a machine only needs to have very limited domain knowledge, not human level intelligence (a.k.a. artificial general intelligence). Think about driving a car. You aren’t using 100% of your total intelligence to drive a car. A large portion is thinking about other things, like disagreeing with this article, singing along with the radio, and probably texting. An autonomous driving system only needs to be capable of processing image data, communicating with computers from other devices related to driving, like other vehicles, traffic signals, and maybe even the road itself, making dynamic calculations based on those data inputs and turning those calculations into actions performed by the vehicle. Any incremental intelligence not related to those core functions is irrelevant for an autonomous driving system.

The magnitude of the technological change is also significantly different in this current wave of advancements in AI and robotics. This wave is more akin to the advent of the farm when humans were still gatherers, or the advent of the factory when we were still farmers. Farms not only organized the production of food, but also encouraged the development of community and trade. Factories organized the production of all goods, encouraged the development of cities, and enabled our modern economic system by institutionalizing the trade of labor for wages. Automation will result in equivalent fundamental changes to the philosophy of production by taking it out of the hands of humans. This could result in societal changes of greater freedom of location and a basic income. In a way, the Automation Age may be an enhanced return to the hunter/gatherer period of humanity where basic needs were provided, originally by nature, in the future by machines. Except in the Automation Age, our purpose will be to explore what it means to be human instead of simply survive.

Disclaimer: We actively write about the themes in which we invest: artificial intelligence, robotics, virtual reality, and augmented reality. From time to time, we will write about companies that are in our portfolio. Content on this site including opinions on specific themes in technology, market estimates, and estimates and commentary regarding publicly traded or private companies is not intended for use in making investment decisions. We hold no obligation to update any of our projections. We express no warranties about any estimates or opinions we make.

Bad Culture Doesn’t Scale

The most important lesson from Uber’s travails is that bad culture doesn’t scale. Talented teams with bad culture can build fantastic businesses, but not businesses that last. A unicorn with bad culture is a unicorn with a bomb strapped to its back — it’s only a matter of time before bad culture catches up and forces disruptive change. Sometimes bad culture rears its ugly head quickly, as it did with Zenefits. Sometimes it doesn’t happen until after multi-billion dollar per year business is established, as it did with Uber.

The culture at Uber wasn’t a secret. It had always been known as an aggressive one, and that culture deserves some credit for helping Uber transform the ride hailing industry; however, the bigger and more established a company becomes, the harder it is to maintain bad culture. Rumors spread, lawsuits happen, and good hires leave because it wasn’t what they signed up for. The media will report every painstaking detail. Advanced companies like Uber also face public backlash from customers, impacting revenue. If Uber were a publicly traded company, the stock would be down at least 30% in the past month given the CEO turmoil. Maybe down 50% for the year adding in the Google lawsuit and other well-publicized troubles.

During our time as public equity analysts, we’ve had the opportunity to cover some great, lasting companies like Apple, Amazon, Google, and Facebook. A common thread between all four of those companies is great culture. When Steve Jobs passed away, we wrote that his greatest achievement wasn’t the iPhone, the iPod, or the Mac, but Apple itself. He left behind a culture of good people driving revolutionary innovation. That might sound simple, but not compromising on your values and consistently hiring the right people that share those values is hard. It’s especially hard for a startup trying to build quickly while bearing the pressure of venture investor expectations.

It’s hard to determine the long-term fallout of Uber’s culture problem. The company has “verbed” itself, much like Google, which allows it a significant brand advantage. One of our teammates has joked that he would, “uber us a Lyft.” With broad leadership change, including the departure of its CEO, Uber has a chance to grow new roots and overcome the negative culture that’s now detracting far more than it ever added.

Disclaimer: We actively write about the themes in which we invest: artificial intelligence, robotics, virtual reality, and augmented reality. From time to time, we will write about companies that are in our portfolio. Content on this site including opinions on specific themes in technology, market estimates, and estimates and commentary regarding publicly traded or private companies is not intended for use in making investment decisions. We hold no obligation to update any of our projections. We express no warranties about any estimates or opinions we make.

Eric Schmidt is Wrong About Automation

At the Viva Tech conference in Paris, Google Chairman Eric Schmidt stated that he believes automation will create more jobs, not eliminate them. I think he’s wrong, and I hope he’s wrong. Disagreeing with the Chairman of the most advanced AI company in the world about automation is a dangerous game, but there are three things that can challenge his statement: timing, incentives, and economic realities. Let’s discuss his position through each of those lenses.

Timing. While Schmidt said he expects AI to create more jobs instead of eliminate them, it’s unclear what time frame he’s considering. When people talk about AI eliminating jobs, it’s almost always on how far out they’re looking into the future. Sometimes that’s five years, sometimes it’s 20, and sometimes it’s 50. At the same conference, GE CEO Jeff Immelt said the idea that robots would run factories in five years is “bullshit.” That I can agree with. The five-year picture of automation isn’t going to result in mass job loss, but the transition will start. Low-skill blue collar jobs will see continued automation. We should start to see autonomous vehicles and continued industrial automation.

Long term, automation isn’t bullshit. If we don’t have machines and software capable of performing most of the tasks we call labor in 30, 40, 50 years, then it will be a failure of Google and our technology ecosystem. We already have machines that can see and hear. We have machines that can roughly manipulate objects in the real world. We have machines that can “understand” enough at a base level to be useful at specialized tasks. Robots don’t get tired, they don’t need breaks, and they don’t get distracted. They will eventually be able to do things with greater precision and sophistication than humans, whether physical work or knowledge work. When robots get sick (broken), they’re much easier to fix or even replace. Robots don’t need to commute to jobs, which saves on energy costs. Robots don’t need paid vacation or catered lunches. For all these reasons, robots will eventually be the most competitive option for the majority of jobs. A few more decades of improvement on artificial intelligence and robotics should yield far more capable machines that can perform almost all work more effectively and more efficiently than humans.

Incentives. Any discussion around job creation or loss is highly political, which means it’s also highly emotional. Many people get scared or even angry when confronted with the possibility of mass automation and human “unemployment”. Eric Schmidt is a savvy politician, and Google is the world’s leader in artificial intelligence development. He doesn’t want the world associating Google with job loss because it could negatively impact their business. He may also have other desires to serve in public office and is setting the stage for those ambitions. Either way, he’s incentivized to be an automation unemployment denier.

But the issue of incentives flows similarly to all executives and CEOs, Immelt included. They’re in an impossible political situation. If Immelt, or any CEO, were to embrace robots capable of eliminating human jobs, there would be backlash among their employees. No worker would be happy to be viewed as a stop gap to automation. There would also be massive PR backlash.

This psychological reality of having to avoid these negative incentives can have real impact on the progress of automation adoption. Decision makers who are caught between looking for improved productivity from automation and maintaining jobs may be forced to seek suboptimal solutions to keep humans employed. Over the past several decades, automation in factories hasn’t meaningfully improved productivity. One reason may be that robot installations approved by executives are done so with the goal of sustaining human jobs rather than maximizing total productivity.

Economic Realities. During his speech, Schmidt argued that not only could automation help create more jobs, but also raise wages. He said that if you “make people smarter” via computers, their wages should increase. In a vacuum this might be true, but in reality this seems to ignore supply and demand.

Let’s take an example in knowledge work automation. In the future, computers are going to be better accountants, financial advisors, actuaries, claims adjusters, etc. than humans. All of these are highly logical jobs driven by information. Per Schmidt’s argument, automating these functions should lead to more jobs, which may be true. Knowledge work jobs might still need a human front end to present the end results from the computer with a human touch (empathy); however, now you’re employing a good customer service operator, not an accountant. In fact, the human might not need to have much more than an entry level understanding of accounting and a positive demeanor to present the script that the computer provides them, which means that far more people are qualified to perform the job of presenting accounting results from a machine than are qualified to be an accountant. Given this great supply of potential workers, it’s hard to see how wages for workers replacing accountants would rival those of accountants today.

Knowledge work seems most sensitive to the economic realities of making humans smarter, but the same can apply to blue collar work. If autonomous trucking becomes a reality in the next decade, that might result in lower freight fees, which might result in increased demand for freight services, which might result in increased demand for dock workers or truck loaders; however, those skills are widely available and now there are a pool of unemployed truck drivers to fill those spots. Again, supply and demand would suggest businesses need not pay higher wages to attract workers to lower-skilled labor.

What We Can Agree On. Instead of debating whether automation will create jobs or eliminate them, we should instead start to set aside the modern dogma that humans must have jobs to survive and be productive. We should consider what a world would look like where humans don’t have to work. A world where all humans don’t need to worry about basic needs because of automation, freeing them to explore what it means to be human. Free to provide value through empathy, community, and creativity — the things robots cannot do.  Maybe we won’t agree on the outcome of automation, but one point we hopefully all agree on is that the future is bright because of automation, not in spite of it.

Disclaimer: We actively write about the themes in which we invest: artificial intelligence, robotics, virtual reality, and augmented reality. From time to time, we will write about companies that are in our portfolio. Content on this site including opinions on specific themes in technology, market estimates, and estimates and commentary regarding publicly traded or private companies is not intended for use in making investment decisions. We hold no obligation to update any of our projections. We express no warranties about any estimates or opinions we make.

Building a Startup Is Like Launching a Rocket

I read about 80 books a year and am always looking for new suggestions from people in technology (email us your recommendations). A number of VCs and technologists recommended Failure Is Not An Option by Gene Kranz, flight controller for the legendary Apollo 13 space mission. After reading the (audio)book, I see four reasons why the book is so beloved among the tech set.

1. Parallels of Launching a Rocket vs a Startup

“Going to the moon was more art than science because they were doing something that had never been done before.”

To boldly go where no man or woman has gone before was the mission of the US space program in the 70s and is the mission for all great startups today. Kranz retells stories of many space missions and preparations that faced calamitous error because they were doing something that had never been done before, but flight control and the astronauts persevered and found solutions. Startups also face frequent, calamitous problems because they too are building a business that does not exist, and they too must persevere. Most startups (aside from SpaceX) aren’t launching rockets into orbit, but they should treat their mission just as seriously.

Read More

The Future of Retail

It’s no secret that online retail is slowly killing offline retail.  In Q4 2016, 8.3% of total US retail sales were online (about $103 billion), up from 5.1% just five years ago (about $53 billion). Offline sales were 91.7% of the total, about $1.1 trillion. We don’t typically talk about the percentage sales that happen offline, but it’s powerful to see how large that market remains. The longer-term question is: how much of total retail will eventually happen online?  We looked at the breakdown of US retail sales by category excluding gas and restaurant expenditures. Based on our analysis, we believe that 55% of total retail sales will eventually happen online, leaving 45% of retail sales for the offline world. But how will brick & mortar retail defend its territory?

We believe the answer lies largely in a combination of artificial intelligence and robotics.  Where AI and robots are superior to humans in terms of efficiency, logic, and raw productivity, we believe humans will remain superior at creativity, community, and experience. Machine-driven retailers are uniquely qualified for convenience, speed and selection. Human-driven retailers are uniquely qualified to create personalized service based on empathy.

Human retailers are uniquely qualified to create personalized service based on mutual understanding – empathy.

The degree to which retailers are successful in leveraging creativity, community and experiences in their stores is the degree to which they will be successful in defending their businesses against online commerce and automated retail.

Given that backdrop, we see the future of retail delivered in three ways:

  • Online Shopping
  • Automated brick & mortar
  • Empathic offline retail

The Future of Online Shopping. Our analysis of US retail sales by category leads us to believe that 55% of total retail sales will eventually happen online. The consumer benefits of convenience, quick shipping and expansive product selection are too powerful to slow the gains that online shopping is enjoying at traditional retail’s expense. Amazon gets it, and they’re playing the long game, aggressively denying short term gains to establish itself as the owner of the operating system for commerce in the future. But Amazon also gets the fact that not all retail is best suited for the internet, which is why we’ve seen them dabbling in automated brick & mortar concepts. More on this below.

More immersive buying experiences will be a major driver of further gains for online shopping. Specifically, augmented reality and virtual reality will allow shoppers to experience a product in lifelike ways before they purchase it. Test out a new outfit in VR and get feedback from your friends. Show your significant other the new couch in your living room with AR before you order custom furniture. The likelihood of returns goes down, customer satisfaction goes up, and so too does the share of online retail.

The Future of Automated Brick & Mortar. We also expect a portion of retail to move to an automated model with few if any employees.  Stores will be monitored by computer vision systems. Shelves will be stocked by robots. Customers will be helped by service robots that understand natural language.  Checkout may resemble Amazon Go where customers simply walk out with their purchases. We’ll likely see the lines between online shopping and automated brick & mortar blur as some stores become more like warehouses for delivery personnel or delivery robots.

Read More