Curated Insights 2019.10.11

Three big things: The most important forces shaping the world

Lower births are a global phenomenon, particularly in the developed world. And while America ages and population growth slows, the rest of the world’s major economies turn into a Florida retirement community and population growth in many cases is on track to turn negative.

When people talk about what nation will own the next century they point to leadership in AI and Machine Learning, where China looks so competitive. But it’s staggeringly hard to grow an economy when you lose a fifth of your working-age population in a single generation. China could invent something as big as the next internet, but when mixed with its demographics have an economy that muddles along. Europe, Japan, and South Korea are the same or worse.

Demographics will slow America’s economy, but they’re a five-alarm fire for other countries. So even assuming equal levels of productivity growth, the U.S. is head and shoulders better off than other developed nations, just given its demographics alone. America could drop the ball on technology while China/Europe/Japan make all the right moves, and America could still remain a much larger and more powerful economy.

TechCrunch founder Michael Arrington recently wrote: “I thought Twitter was driving us apart, but I’m slowly starting to think half of you always hated the other half but never knew it until Twitter.” This is a good point that highlights something easy to overlook: 1) everyone belongs to a tribe, 2) those tribes sometimes fundamentally disagree with one another, 3) that’s fine if those tribes keep their distance, 4) the internet increasingly assures that they don’t. Opening your mind to different perspectives is good and necessary. But when fundamental, unshakable views that used to be contained within tribes expose themselves to different tribes, people become shocked to learn that what’s sacred to them isn’t always a universal truth. The range of political opinions has always been extreme, but what we’ve seen over the last decade is what happens when the warm blanket of ideological ignorance is removed.

The best economic news no one wants to talk about

So, let’s play a game of wish-casting. Imagine a world where wage growth was truly stagnant only for workers in high-wage industries, such as medicine and consulting. Imagine a labor market where earnings growth for low-wage workers, such as those who work in retail and restaurants, had doubled in the past five years. Imagine an economy where wages for the poorest Americans were rising twice as fast as hourly earnings for high-wage earners. It turns out that all three of those things are happening right now.

One reason you haven’t heard this economic narrative may be that it’s inconvenient for members of both political parties to talk about, especially at a time when economic analysis has, like everything else, become a proxy for political orientation. For Democrats, the idea that low-income workers could be benefiting from a 2019 economy feels dangerously close to giving the president credit for something. This isn’t just poor motivated reasoning; it also attributes way too much power to the American president, who exerts very little control over the domestic economy. Meanwhile, corporate-friendly outlets, such as The Wall Street Journal’s editorial pages, have reported on this phenomenon. But they’ve used it as an opportunity to take a shot at “the slow-growth Obama years” rather than a way to argue for the extraordinary benefits of tight labor markets for the poor, much less for the virtues of minimum-wage laws.

Democrats don’t want to talk about low-income wage growth, because it feels too close to saying, “Good things can happen while Trump is president”; and Republicans don’t want to talk about the reason behind it, because it’s dangerously close to saying, “Our singular fixation with corporate-tax rates is foolish and Keynes was right.”

But good things can happen while Trump is president, and Keynes was right. “Tighter labor markets sure are good for workers who work in low-wage industries,” Bunker told me. “This recovery has not been spectacular. But if we let the labor market get stronger for a long time, you will see these results.”

Charles Schwab and the new broker wars

Schwab now derives more than half of its revenue from net interest income, and the company estimates that it will lose $75 million to $150 million in revenue for every quarter-point cut by the Federal Reserve. If we get four more cuts over the next 12 months, Schwab could lose $600 million, about 6% of its estimated $10.6 billion total.

“People underestimate how much the economics of Schwab’s business comes from investing client cash,” says Steven Chubak, an analyst with Wolfe Research. “Rising rates were a very good story for them, but rates may now be going in the other direction, and that will create headwinds,” says Devin Ryan, an analyst with JMP Securities.

Schwab can withstand the revenue loss. It is one of the most broadly diversified brokerages, including asset-management, custodial, and back-office services for institutional investors. Based in San Francisco, the firm oversees $3.7 trillion in client assets, including $1.55 trillion custodied by registered investment advisor firms, or RIAs. Schwab is the largest RIA custodian in the country. The company sponsors mutual funds and exchange-traded funds. Its Intelligent Portfolios service—automated managed accounts of ETFs—has grown into the largest robo-advisor with $30 billion in assets.

The big profit center for Schwab is now its bank. With more than $276 billion in assets, Schwab Bank is larger than Ally Financial, KeyCorp, and Fifth Third Bancorp, according to S&P Global Market Intelligence. Schwab Bank recently crossed a regulatory threshold, subjecting it to stiffer federal stress-test, capital, and liquidity requirements.

As rates increased in recent years, Schwab Bank became the tail that wags the company dog. Net interest revenue from the bank amounted to $5.8 billion, or 57% of Schwab’s total revenue of $10.1 billion in 2018, up 36% year over year. Management and administrative fees were 32% of revenue in 2018, with trading and related revenue rounding out the pie.

Schwab’s revenue base now looks well balanced between RIA sources (such as custodial fees) and retail brokerage, Chubak says. The company’s low-cost ETFs and robo-advisory service are marketplace winners. Schwab recently rolled out a premium subscription advisory service offering “unlimited guidance” for $30 a month and a one-time planning fee of $300. Schwab CEO Bettinger said on a recent call with analysts that the premium subscription service “seems to have really taken off in terms of client interest and response.”

Schwab’s platform for RIAs is considered one of the strongest suites of tools and software in the industry. And it is benefiting as advisors break away from Wall Street brokerage houses. The trend has been going on for a decade, but it may be gaining momentum. Bettinger said recently that the breakaway RIA trend began “to pick up a bit again in the second quarter.” Schwab recently launched an upgraded version of its portfolio-management software to compete more effectively.

“Schwab’s competitive strength is their enormous stronghold on the advisor community,” says Thomas Peterffy, chairman of Interactive Brokers. “They have cultivated that for years.” The RIA industry is also consolidating into firms with 50 to 100 advisors, says Chip Roame, managing partner of Tiburon Strategic Advisors, a financial consulting firm. That’s good news for Schwab, since larger firms with more trading, analytics, and custodial requirements are likely to bring assets to the firm.

The tech stock that Apple, AMD, and Nvidia can’t do without

TSMC’s client list includes the world’s top technology companies, such as Apple (AAPL), Qualcomm (QCOM), Huawei Technologies, Nvidia (NVDA), and Advanced Micro Devices (AMD). They all rely on TSMC to make the most demanding chips used in smartphones, servers, artificial intelligence applications, and networking devices.

JPMorgan estimates that TSMC accounts for about 50% of the world’s foundry revenues and 80% to 90% of the industry’s profits.

TSMC has consolidated its market share in recent years because its foundries were the first to offer 7-nanometer chip production at significant volume. Smaller chips offer greater performance and improved power efficiency. The entire chip industry is rapidly trying to get to a 7nm (and lower) manufacturing process, but most manufacturers have yet to make the transition. Intel (INTC), which fabricates most of its own chips, is unlikely to have 7nm products before 2021.

Curated Insights 2019.09.27

Why these two innovations in artificial intelligence are so important: Eye on A.I.

The first item was news that a Hong Kong-based biotechnology startup, InSilico Medicine, working with researchers from the University of Toronto, had used machine learning to create a potential new drug to prevent tissue scarring. What’s eye-popping here is the timescale: just 46 days from molecular design to animal testing in mice. Considering that, on average, it takes more than a decade and costs $350 million to $2.7 billion to bring a new drug to market, depending on which study one believes, the potential impact on the pharmaceutical industry is huge.

What’s also interesting here is that InSilico used reinforcement learning, an A.I. technique that hasn’t yet impacted business much. Reinforcement learning is notable because it doesn’t require the vast pools of structured, historical data that other A.I. methods do. Here researchers used reinforcement learning to rapidly design 30,000 new molecules and then narrow them down to six, which were synthesized and further tested in the lab. Look for more A.I. breakthroughs like this to start upending the balance of power between biotech startups and Big Pharma.

The second piece of underappreciated news is that researchers at DeepMind, the London A.I. shop owned by Google parent Alphabet, and Imperial College London, successfully used a deep neural network to find more precise answers to quantum mechanical problems. That’s basically the physics that underpins all of chemistry.

To date, the only element for which we can completely solve the underlying quantum equations is the simplest, hydrogen, which has just one proton and one electron. For every other element, we rely on approximations. Get better approximations, and you potentially get new chemistry – and that means new materials. Think room temperature superconductors or new kinds of batteries that will vastly extend the range of electric vehicles. DeepMind’s A.I.-powered approximations were in some cases almost an order of magnitude better than previous methods. If you’re Dow or DuPont, or Formosa Plastics or LG Chem, that sort of advantage could be worth billions.

Why prescription drugs cost so much more in America

All over the world, drugmakers are granted time-limited monopolies — in the form of patents — to encourage innovation. But America is one of the only countries that does not combine this carrot with the stick of price controls. The US government’s refusal to negotiate prices has contributed to spiralling healthcare costs which, said billionaire investor Warren Buffett last year, act “as a hungry tapeworm on the American economy”. Medical bills are the primary reason why Americans go bankrupt. Employers foot much of the bill for the majority of health-insurance plans for working-age adults, creating a huge cost for business.

Other drugs are more innovative — and their development undeniably expensive. According to Tufts University, the average is $2.6bn per drug, up 145 per cent in the past 10 years. Most drug candidates fail; those that do make it to later stages must go through expensive clinical trials. In support of the drug companies’ argument, one 2015 study found that for every extra $2.5bn a company made in sales, it produced one extra drug.

Robot pilot that can grab the flight controls gets its plane licence

A recent conversion of US military F-16 fighter jets into drones cost more than a million dollars each. ROBOpilot can be inserted into any aircraft and just as easily removed afterwards to return it to human-controlled operation.

“It looks like an impressive achievement in terms of robotics,” says Louise Dennis at the University of Liverpool. “Unlike an autopilot which has direct access to the controls and sensors, the robot is in the place of a human pilot and has to physically work the controls and reads the dials.” The makers suggest that ROBOpilot will be useful for tasks including transporting cargo, “entry into hazardous environments”, and intelligence, surveillance and reconnaissance missions.

Entrepreneurs hope microbes hold the key to a food revolution
A taxonomy of moats

Curated Insights 2019.08.09

Good for Google, bad for America

A.I.’s military power is the simple reason that the recent behavior of America’s leading software company, Google — starting an A.I. lab in China while ending an A.I. contract with the Pentagon — is shocking. As President Barack Obama’s defense secretary Ash Carter pointed out last month, “If you’re working in China, you don’t know whether you’re working on a project for the military or not.”

Netflix is not a tech company

Hence, Netflix isn’t using TV to leverage some other business – TV is the business. It’s a TV company. Amazon is using content as a way to leverage its subscription service, Prime, in much the same way to telcos buying cable companies or doing IPTV – it’s a way to stop churn. Amazon is using Lord of the Rings as leverage to get you to buy toilet paper through Prime. But Facebook and Google are not device businesses or subscription businesses. Facebook or Google won’t say ‘don’t cancel your subscription because you’ll lose this TV show’ – there is no subscription. That means the strategic value of TV or music is marginal – it’s marketing, not a lock-in.

Apple’s position in TV today is ambivalent. You can argue that the iPhone is a subscription business (spend $30 a month and get a phone every two years), and it certainly thinks about retention and renewals. The service subscriptions that it’s created recently (news, music, games) are all both incremental revenue leveraging a base of 1bn users and ways to lock those users in. But the only important question for the upcoming ‘TV Plus’ is whether Apple plans to spend $1bn a year buying content from people in LA, and produce another nice incremental service with some marketing and retention value, or spend $15bn buying content from people in LA, to take on Netflix. But of course, that’s a TV question, not a tech question.

Why we sold Trupanion

Why is Trupanion not outpacing the industry when it is the first name many pet owners hear? It could be that pet owners hear about pet insurance from their vet, go home, compare prices, and choose a more affordable option. It’s not an impossible problem for Trupanion to solve, but again, consumers don’t typically understand insurance value until they file a claim.

But at an investor event we attended a few months ago, Darryl said that he was making plans to switch to an executive chairman role in 2025. While we understood Darryl’s choice on a personal level, it also sharply increased our uncertainty around whether company management will be able to successfully build a moat and transform pet insurance as we had hoped. In our opinion, Trupanion will continue to need a visionary leader in the CEO role and finding another visionary to replace Rawlings will be a massive challenge.

Given the industry’s rapid growth, we think it’s perfectly normal for both regulators and pet insurance companies to have some growing pains. While Trupanion has been fined, faces more state investigations, and admits it should have paid more attention to regulators as a stakeholder, we considered these matters minor to our thesis. Regulators may require Trupanion’s Territory Partners to be licensed in all states, but this is more like a speed bump rather than a roadblock.


TGV Intrinsic on MercadoLibre

Network effects are among the highest entry barriers for competitors to build and leverage business. As market leader, MercadoLibre has been able to permanently focus on strengthening the network effects of the marketplace and eliminate points of conflict in transaction processing between buyer and seller. The most important point of conflict between seller and buyer in the past were the payment arrangements. To simplify this process, MercadoLibre launched its own payment service “MercadoPago” in 2004 (comparable to PayPal). MercadoPago provides a secure way to pay for goods and simplified the coordination between buyers and sellers in terms of payment. Today, over 90% of the value god goods sold on MercadoLibre is paid with MercadoPago, which equates to a payment volume of 11 billion US dollars.

Apart from that, logistics costs in many Latin American countries pose a major hurdle for buyers and sellers. The investments required to setup one’s own logistics system are high, delivery times in Latin America are relatively long, and service is rather mediocre. With the founding of MercadoEnvios in 2013, MercadoLibre took over an ever more extensive control over the logistics of goods in several steps. Today, MercadoEnvios operates its own logistics centres, takes over the first mile from the seller or organises the last mile to the end customer with selected partners. In 2018, at least part of the logistics was taken over by MercadoEnvious for 66% of the goods sold through the marketplace. Thanks to the sizeable investments in a proprietary payment system and the continuous expansion of its own logistics, MercadoLibre has massively expanded its marketplace and the network effect that has been set in motion over the past two decades. The value of these investments is reflected in the growth in the number of transactions amounting to 28% per annum over the past decade.

The value of this ecosystem lies in the ever-growing economy of scale. MercadoLibre has more touch points with its customers than its competitors, be it specialised online retailers, payment service providers, or logistics companies. At each of these points of contact, MercadoLibre can distribute its costs in customer acquisition to more services than its competitors. As a result, the costs for new customers per product are lower than for competitors. Second, MercadoLibre can freely decide which areas of a customer relationship to monetise and which not. For example, MercadoLibre may offer MercadoPago payment service to new brick-and-mortar retailers for free but would require a marketplace transaction fee for this merchant’s online product sales. A specialised payment provider does not have this flexibility. As a result, MercadoLibre has created a flexible and cost-effective customer acquisition engine that only very few companies have.


Zebras can change their stripes

Since 2012, JUVE has gone on to secure many other high profile “free” transfers with established winners such as Dani Alves and Sami Khediera, but then also find those that have significant upside potential like Adrien Rabiot, Kingsley Coman, and Emre Can. By far, the most impactful and value-accretive “free” transfer was Paul Pogba. In 2012, Juventus signed 19 year-old Pogba from Manchester United, and only four seasons later (and after winning four league titles) sold him back to Manchester United for a world-record fee of $116 million. In total, since 2011, the top 10 free transfer signings by Juventus have created $204m of value (i.e. market value of the players at time of signing) and over $135 million of cash from transfer sales proceeds. Yes, Pogba was an outlier, but JUVE has utilized the free transfer market better than any other club over the past decade.

Let us not forget this is a business, and Ronaldo prints money. Before the ink dried on the contract, the Ronaldo effect took Juventus, and Italy, by storm. His name generated over $60 million in jersey sales in one day – that is the best global branding a club can ask for. JUVE’s Twitter account showed a 10% increase in followers on the day he was signed. Then ESPN acquired the U.S. Serie A TV rights at a massive step-change in the fee ($55 million per year, versus $28 million previously) only one month later. And to no one’s surprise, the first game aired on August 18th showcasing Ronaldo in his new black and white jersey. Your author duly signed up for ESPN+ exclusively to live stream I Bianconeri. The acquisition led to a nice bump in GreenWood’s performance, and most importantly, Ronaldo helped his team win another Serie A title.

The periodic table of investments

Winner-take-all phenomenon rules the stock market, too

Just 1.3% of the world’s public companies account for all the market gains during the past three decades. Outside the U.S., the gains are even more concentrated, with less than 1% of all equities driving all of the net appreciation in share prices.

Just five companies — Apple Inc., Microsoft Corp., Amazon.Com Inc., Alphabet Inc. (Google) and Exxon Mobil Corp. — accounted for 8.3% of global net wealth creation. It is hard to imagine a greater example of the winner-take-all distribution — these five companies account for just 0.008% of the total sample set of 62,000 publicly traded companies. Expand that to the top 0.5%, or 306 companies, and they account for 73% of global net wealth creation. The best performing 811 companies (1.33% of the total) accounted for all net global wealth creation.


Negative rates could happen in America, too

What’s behind negative interest rates? Many observers blame central banks like the European Central Bank (ECB) and the Bank of Japan (BOJ) that are taxing banks’ excess reserves with negative deposit rates and have made bonds scarcer by removing them from the market through their purchase programs. The BOJ now owns about half and the ECB about 30% of the bonds issued by their respective governments, according to Bloomberg.

However, we believe central banks are not the villains but rather the victims of deeper fundamental drivers behind low and negative interest rates. The two most important secular drivers are demographics and technology. Rising life expectancy increases desired saving while new technologies are capital-saving and are becoming cheaper – and thus reduce ex ante demand for investment. The resulting savings glut tends to push the “natural” rate of interest lower and lower.

LBOs make (more) companies go bankrupt, research shows

According to researchers at California Polytechnic State University, roughly 20 percent of large companies acquired through leveraged buyouts go bankrupt within ten years, as compared to a control group’s bankruptcy rate of 2 percent during the same time period.


Is great information good enough? Evidence from physicians as patients

We compare the care received by a group of patients that should have the best possible information on health care service efficacy—i.e., physicians as patients—with a comparable group of non-physician patients, taking various steps to account for unobservable differences between the two groups. Our results suggest that physicians do only slightly better in adhering to both low- and high-value care guidelines than non-physicians – but not by much and not always.


Health facts aren’t enough. Should persuasion become a priority?

Those who were most opposed to genetically modified foods believed they were the most knowledgeable about this issue, yet scored the lowest on actual tests of scientific knowledge. In other words, those with the least understanding of science had the most science-opposed views, but thought they knew the most.

Curated Insights 2019.06.28

Facebook, Libra, and the long game

And this is when this bet would pay off for Facebook (and the second point I missed in my earlier analysis): the implication that digital currencies will do for money what the Internet did for information is that the very long-term trend will be towards centralization around Aggregators. When there is no friction, control shifts from gatekeepers controlling supply to Aggregators controlling demand. To that end, by pioneering Libra, building what will almost certainly be the first wallet for the currency, and bringing to bear its unmatched network for facilitating payments, Facebook is betting it will offer the best experience for digital currency flows, giving it power not by controlling Libra but rather by controlling the most users of Libra.

Forget the mall, shoppers are buying Gucci at airports

For the first time last year, Estée Lauder Co. generated more revenue at airports globally than at U.S. department stores, which for decades had been beauty companies’ biggest sales driver … “Very few channels have almost guaranteed traffic,” said Olivier Bottrie, who heads Estée Lauder’s global travel-retail business. “When a department store goes away, it’s not a major catastrophe. But if a major airport went away, it would be a major catastrophe.”

Employees who stay in companies longer than two years get paid 50% less

Staying employed at the same company for over two years on average is going to make you earn less over your lifetime by about 50% or more …

In 2014, the average employee is going to earn less than a 1% raise and there is very little that we can do to change management’s decision. But, we can decide whether we want to stay at a company that is going to give us a raise for less than 1%. The average raise an employee receives for leaving is between a 10% to 20% increase in salary.

Curated Insights 2019.01.04

The customer acquisition pricing parade

“One spectator, determined to get a better view, stands on their tiptoes. It works well initially until everyone else does the same. Then, the taxing effort of standing on your toes becomes table stakes to be able to see anything at all. Now, not only is any advantage squandered, but we’re all worse off than we were when we first started.”

“Marketing is increasingly cheap. Trust is increasingly expensive. Attracting eyeballs no longer sets you apart. Building trust among those who have their eyes on you, does. Getting people’s attention is no longer a skill. Keeping people’s attention is.”

To decrease spending and increase profitability, the holding companies of tomorrow will shift their attention from controlling supply to controlling demand — from building around industries to building around audiences.  

Re-marketing to an existing customer is significantly cheaper than trying to persuade a first time customer to buy your product — sometimes nearly 90% cheaper.

Companies who cater to the needs of passionate customers will benefit from lowered customer acquisition costs and higher lifetime value (LTV), reduced churn and increased loyalty. Once a paying customer is acquired, companies can cross-sell and up-sell them into different products, categories, and even brands. The fight to find that customer will be much easier leading to an increase in transaction volume. As they reduce friction in the payment process and increase customer loyalty, they’ll accrue data behind customer cohorts leading to a customer-centric experience.

Companies who cater to their customers and develop direct relationships with them, will own the future.

Working more magic at Disney

Walt Disney has in a sense become more Disney-like in how it earns its profits. After 20 years of being dominated by television, especially cable, the company is returning to its roots in films and theme parks. Seven years ago, for each $1 in operating profit that Disney made from its parks and studios, it generated $3 in TV. During the fiscal year ended September, parks and studios retook the lead.

The coming streaming platform will be reported to Wall Street as a separate Netflix-like division within Disney. Investors will see how much cash the unit is paying for content. New Disney, so to speak, will pay this money to Old Disney. “What I’ve discovered is, businesses in a traditional space that want to innovate and spend money to do so, they park the cost of innovation in their traditional businesses,” Iger says. “Those businesses all kind of suffer from the cost of that innovation, because it’s not typically monetized right away. You can get impatient to the point of losing interest and abandoning innovation, because you don’t have the patience to wait for it to really pay off.”

For decades, Disney was largely a moviemaker with theme parks, although television has long been part of the mix; The Mickey Mouse Club, which began on ABC in 1955, helped finance Disneyland, and the Disney Channel has been a cable mainstay since the 1980s. But in 1995, Disney surprised investors with a $19 billion acquisition of Capital Cities/ABC, gaining a prosperous TV network and a thriving ESPN. As cable spread service across the nation, and TV producers learned how to extract higher fees from cable operators for their content, the small screen became Disney’s big earner.

Iger planted the roots of Disney’s growth spurt in its traditional businesses when he rolled up major story-telling outfits that weren’t for sale. He did that by visiting their bosses one-on-one: Steve Jobs, culminating in the $7.4 billion purchase of Pixar in 2006; Isaac Perlmutter, for a $4 billion acquisition of Marvel Entertainment in 2009; and George Lucas, in a $4 billion deal for Lucasfilm in 2012. What has followed has been a film boom for the ages.

This year, Disney will again become the only studio in history to reap $7 billion in worldwide box office receipts: $4 billion internationally and $3 billion in the U.S. It also did so in 2016. And Disney makes eight to 10 films a year; some big studios make two dozen.

Box office results relate to return on invested capital the way dunking a basketball relates to winning a game: The former gets the crowd’s attention, but people with stakes care mostly about the latter. The broader film industry operates with a single-digit return, Iger says. Yet there are years when Disney’s film returns top 30%. As a result, studio operating income has multiplied more than fourfold since 2011, to nearly $3 billion during the fiscal year ended in September.

Does AI make strong tech companies stronger?

First, though you need a lot of data for machine learning, the data you use is very specific to the problem that you’re trying to solve. GE has lots of telemetry data from gas turbines, Google has lots of search data, and Amex has lots of credit card fraud data. You can’t use the turbine data as examples to spot fraudulent transactions, and you can’t use web searches to spot gas turbines that are about to fail. That is, ML is a generalizable technology – you can use it for fraud detection or face recognition – but applications that you build with it are not generalized. Each thing you build can only do one thing. This is much the same as all previous waves of automation: just as a washing machine can only wash clothes and not wash dishes or cook a meal, and a chess program cannot do your taxes, a machine learning translation system cannot recognise cats. Both the applications you build and the data sets you need are very specific to the task that you’re trying to solve (though again, this is a moving target and there is research to try to make learning more transferable across different data sets).

So: as an industrial company, do you keep your own data and build the ML systems to analyse it (or pay a contractor do do this for you)? Do you buy a finished product from a vendor that’s already trained on other people’s data? Do you co-mingle your data into that, or into the training derived from it? Does the vendor even need your data or do they already have enough? The answer will be different in different parts of your business, in different industries and for different use cases.

This takes me to a metaphor I’ve used elsewhere – we should compare machine learning to SQL. It’s an important building block that allowed new and important things, and will be part of everything. If you don’t use it and your competitors do, you will fall behind. Some people will create entirely new companies with this – part of Wal-Mart’s success came from using databases to manage inventory and logistics more efficiently. But today, if you started a retailer and said “…and we’re going to use databases”, that would not make you different or interesting – SQL became part of everything and then disappeared. The same will happen with machine learning.


One giant step for a chess-playing machine

Most unnerving was that AlphaZero seemed to express insight. It played like no computer ever has, intuitively and beautifully, with a romantic, attacking style. It played gambits and took risks. In some games it paralyzed Stockfish and toyed with it. While conducting its attack in Game 10, AlphaZero retreated its queen back into the corner of the board on its own side, far from Stockfish’s king, not normally where an attacking queen should be placed.

Yet this peculiar retreat was venomous: No matter how Stockfish replied, it was doomed. It was almost as if AlphaZero was waiting for Stockfish to realize, after billions of brutish calculations, how hopeless its position truly was, so that the beast could relax and expire peacefully, like a vanquished bull before a matador. Grandmasters had never seen anything like it. AlphaZero had the finesse of a virtuoso and the power of a machine. It was humankind’s first glimpse of an awesome new kind of intelligence.

Tellingly, AlphaZero won by thinking smarter, not faster; it examined only 60 thousand positions a second, compared to 60 million for Stockfish. It was wiser, knowing what to think about and what to ignore. By discovering the principles of chess on its own, AlphaZero developed a style of play that “reflects the truth” about the game rather than “the priorities and prejudices of programmers,” Mr. Kasparov wrote in a commentary accompanying the Science article.

What is frustrating about machine learning, however, is that the algorithms can’t articulate what they’re thinking. We don’t know why they work, so we don’t know if they can be trusted. AlphaZero gives every appearance of having discovered some important principles about chess, but it can’t share that understanding with us. Not yet, at least. As human beings, we want more than answers. We want insight. This is going to be a source of tension in our interactions with computers from now on.

Maybe eventually our lack of insight would no longer bother us. After all, AlphaInfinity could cure all our diseases, solve all our scientific problems and make all our other intellectual trains run on time. We did pretty well without much insight for the first 300,000 years or so of our existence as Homo sapiens. And we’ll have no shortage of memory: we will recall with pride the golden era of human insight, this glorious interlude, a few thousand years long, between our uncomprehending past and our incomprehensible future.


Evaluating early stage startups — The three metrics that matter

Defining “fast growth” depends on stage, but for early (Seed or Series A), growing 100% YoY is typically pretty solid. Paul Graham (PG) famously looks for 5–7% weekly growth for companies in Y Combinator, and his rationale is pretty simple: “a company that grows at 1% a week will grow 1.7x a year, whereas a company that grows at 5% a week will grow 12.6x.” When you consider the compounding effects of this growth, it means a company starting with $1,000 in revenue and growing at 1% will be at $7,900 per month four years later, whereas the company growing 5% per week will be bringing in more than $25 million per month.


The founder’s guide to understanding investors

When we dig deeper, the degree to which early-stage investing is a grand slam business is shocking. First, amongst early stage investors, the returns are disproportionately distributed. The Kauffman Foundation, an investor in many VC funds, found the top 20 VC firms (~3% of VC firms), generate 95% of all venture returns. Second, outside of the top 20 VC firms, most lose money! A study found the top 29 VC firms made a profit of $64B on $21B invested, while the rest of the VC universe lost $75B on $160B invested.

As early-stage investing operates on a power law, Paul Graham (founder of Y Combinator) mentions “You [as an investor] have to ignore the elephant in front of you, the likelihood they’ll [the startup] succeed, and focus instead on the separate and almost invisibly intangible question of whether they’ll succeed really big.” He highlights there are 10,000x variations (!) in startup investing returns, meaning top investors must have the mindset of willing to strike out in order to hit grand slams.

There needs to be room for your startup to capture a large share of this market. Elad Gil (early investor in Airbnb, Coinbase, Gusto, Instacart, Stripe), explains this means i) the market is structurally set up to support multiple winners, but ii) if the market only supports one winner and customers are currently not served well – there is an opportunity to dominate the market.

At the Series A stage, investors are mainly looking to see if PMF is achieved. This evaluation can be qualitative – Marc Andreessen (co-founder of Netscape and Andreessen Horowitz, an early investor in Facebook, Twitter, Wealthfront, Slack) notes, on the inside, “you can always feel product/market fit when it’s happening. The customers are buying the product just as fast as you can make it — or usage is growing just as fast as you can add more servers. Money from customers is piling up in your company checking account. You’re hiring sales and customer support staff as fast as you can. Reporters are calling…”

When investors are evaluating for PMF, Rachleff notes that the best test is to see if the product is growing exponentially with no marketing, meaning the product is so good it grows through word of mouth. Top investors often don’t want to see marketing spend because it shows care for vanity metrics (things that don’t matter) rather than building an amazing product that people engage with (which does matter).

Not all buzzwords will fulfill their potential and result in a disruptive technology shift though. As a founder, you can reduce this risk by avoid starting a startup on that shift until the technology adoption is growing quickly and reaches a multi-hour per day level of usage. Sam Altman expands, “It’s very hard to differentiate between fake trends and real trends…If you think hard and you really pay attention, sometimes you can. The metric I use to differentiate between a real trend and a fake trend is similar to loving a product. It’s when there is a new platform that people are using many hours every day.”

To believe the startup can fulfill grand slam potential, investors want to see the startup has verified their assumptions on how users find the product in a repeatable and scalable manner. This is also called a go-to-market strategy (GTM).

Bill Gurley (major early investor in Uber, Stitch Fix, Zillow, etc.) called a unique GTM the most under-appreciated part about startups. It’s not about who did it first, but who did it right. Gurley looks to see if the startup has two things: (1) An interesting way to get into the market; (2) A way to establish themselves once in the market. The word ‘unique’ is important here. Replicating existing GTM strategies is often too costly because incumbents have already dried up the channel(s) to market and sell to customers. As a founder, you need to find a unique GTM that is repeatable and scalable. The good news here is that if you succeed, you’ll be able to keep out competitors by saturating the new channels.

Andy Rachleff has a second perspective on how startups can avoid competition. With his adaptation to Clayton Christensen’s (Harvard Business School Professor) disruption theory, startups can compete with reduced competition in either two ways. They can compete via new-market disruption – targeting a new set of users and competing on different characteristics (e.g. instead of price, focus on experience) than competitors, or they can compete via low-end disruption – targeting the same set of users as incumbents, but offering a greatly reduced product at a lower price point.

Along with the above quote, Bill Gurley tests if executives at the startup have a notion of insane curiosity – constantly learning new ways to win. To evaluate this, he asks questions on what information (e.g. books, podcasts) executives learn from, how they engage with it, and then probes if they are trying to use that information to majorly improve themselves or their business.

Curious folks tinker. Obsessively curious folks solve the hardest problems that require endless tinkering. If you are obsessively curious and fail with your original plan, odds are you will use your learnings and pivot into a big market that loves their product.

If a founder is obsessively curious, they can navigate the idea maze. By running a founder through the idea maze, investors evaluate if the founder understands all permutations of their idea, why their plan is superior to all other competitors, and which turns to lead to treasure versus which ones lead to certain death. It’s important for a founder to thoroughly know their idea maze, it can save years by not going down the wrong path, in addition to convincing investors you know can be a grand slam.

Rating citizens – can China’s social credit system fix its trust deficit?

In some Chinese cities, if you have wilfully defaulted on paying your debts, callers to your mobile phone will hear this message instead of the usual ringback tone: “This is a friendly reminder from the people’s court of XX city. The person you have just called has been declared a trust-breaker subject to enforcement by the court …”

Given that 80 per cent of respondents in a 2018 opinion poll conducted across China have approved of the social credit system, it seems that most Chinese, for now, do not consider the drastic surveillance scheme a violation of their privacy. Instead, most see the merit of the system in the perks they may enjoy and its potential in fostering trustworthiness in society.

However, the feasibility of this is highly questionable since the government is simultaneously the enforcer, the appraiser and the appraisee. For instance, many local governments in China have often failed to repay debt ranging from a few thousand to tens of million yuan, including loans, payment to contractors, and compensation for seized land. Since local governments are the ones that assess trustworthiness and mete out punishment, will they themselves be subject to the same penalties as other defaulters?

In this county, slandering others online will take 100 points off your social credit rating, while manufacturing and selling fake products will set you back by merely 35 points. Someone who may be rightfully seeking redress by occupying government offices may be slapped with a 50-point deduction, the same penalty as someone who has given or received bribes.

Curated Insights 2018.12.14

The Facebook Fed

I think a good analogy is that Facebook is the Federal Reserve of web publishing. It can turn its dial and blast millions of visitors to numerous publishers, allowing everyone to have more eyeballs to sell to and more rising traffic numbers with which to attract investment. Facebook turning on the traffic fire hose is like loose monetary policy that stimulates the economy for everyone.

Of course, Facebook could tighten policy, pulling traffic (liquidity) and leaving weaker players parched. When the Fed tightens policy, shaky borrowers who depend on ample lending are hit hardest. When Facebook tightens policy, second-tier publishers that totally rely on Facebook are hit hardest.

Digital divide is wider than we think, study says

Over all, Microsoft concluded that 162.8 million people do not use the internet at broadband speeds, while the F.C.C. says broadband is not available to 24.7 million Americans. The discrepancy is particularly stark in rural areas. In Ferry County, for example, Microsoft estimates that only 2 percent of people use broadband service, versus the 100 percent the federal government says have access to the service.

The hardest problem in finance

But here’s the catch, this table assumes you get this rate of return year after year after year. The real world is not so accommodating. The table above shows that you can earn 4% a year for 34 years before running out of money, but let’s look at what happens if a nasty bear market were to arrive as soon as you retire. The chart below shows one way in which an investor can arrive at a 4% CAGR over a 30 year period.

The danger of assuming compound annual growth rates when making long term projections can be seen in the chart below. The black line shows that spending a constant $40k annually, using the returns from the previous chart, an investor would run out of money in the 19th year. Spending 4% and assuming a 2% inflation rate, a more realistic assumption, an investor would run out of money in just 15 years. Side note, if the returns above were to happen in reverse, in other words the bear market comes at the end of the period, an investor with the same spending would be left with $1.3 million.

Curated Insights 2018.12.07

The money and math behind our newsletter headlines

That means that the winning headline results in 4,620 incremental opens per newsletter. Over the course of a month, that is 115,000+ incremental opens. If we apply a conservative estimate based on some of our data from 2015 of how many of those incremental opens try the CB Insights free product (0.09%) and that only 1% of those trials buy a subscription, we’re looking at nearly $625,000 of incremental revenue.

Facebook is undervalued

The result of billions of readers providing free content for each other is a massive network effect and a business that can serve each incremental ad at a very low marginal cost. The company’s 85% gross margin leaves a lot of meat on the bone that the company can spend on hiring new engineers and developing new technology that is needed to maintain and grow the business. But even after paying the engineers, growing the headcount, spending over $9 billion on R&D, and paying the tax bill, shareholders are still left with around 40 cents of profit for each dollar of revenue. Facebook reinvests these earnings into data centers, technology, and the occasional acquisition, but even so, the cash in the till has been piling up, and now exceeds $40 billion and counting ($14 per share). Facebook has spent $10 billion on buybacks in the last year, and this will likely accelerate in the coming years.

Facebook’s business is still growing fast (33% revenue growth last quarter), and despite the company’s large size, the runway is still long. Benedict Evans, a partner at the VC firm Andreessen Horowitz, recently pointed out that roughly $1 trillion is spent each year by businesses who are trying to reach customers who are asking the question: “What should I buy?”. Facebook has roughly 5% of this market, which I think is a share that will inevitably rise over the coming years as advertising dollars continue to shift from traditional media to the much higher ROI advertising platforms like Google, Facebook, and Amazon.

A strong network leads to Facebook grabbing a much greater share of the $1 trillion (and growing) global ad/marketing industry. Revenue from ads alone could easily be twice the current size, which means that Facebook is a $100 billion business before counting any of the possible upside from payments, messaging, or any of the other potential business lines that the company could develop on the back of its large network effect.


Accelerator demand should grow ten-fold according to top 500 supercomputers

If the data center market as a whole were to follow in the footsteps of HPC, accelerators could displace Intel CPUs as the workhorse of enterprise computing.

If processors were to grow to 50% of total server component costs, with accelerators capturing 70% of processors, as is the case in the HPC market today, the market for accelerators could scale from $4 billion1 today to $24 billion. Currently, the size of the computer server market is $80 billion globally, according to IDC. If manufacturers take a 15% cut, then $68 billion accrues to component suppliers such as Intel, Micron, and Nvidia. Today we estimate that processors account for roughly 35% of server component costs, the balance allocated to memory, storage, and networking. With HPC and AI, servers are evolving toward higher compute configurations, pushing processors and accelerators to 80% of total component costs.

As shown by the latest IDC data above, the server market as a whole continues to grow, driven primarily by higher average selling prices (ASPs). With denser configurations, server ASPs could climb from today’s price of $8,000 to $10,000. Even at the current volume of 10 million units per year, that would grow the server market 25% from $80 to $100 billion, pushing the accelerator total addressable market to $35 billion.

Today, the accelerator market is dominated by Nvidia’s data center products, which we estimate will generate $3.2 billion of revenue in 2018. Nvidia estimates that its data center business has a total addressable market of $50 billion—a figure larger than Intel’s data center business, which seems difficult to reconcile. That said, as the latest data from IDC and Top 500 shows, the server market is undergoing a fundamental shift. The decay of Moore’s Law has increased total server spend as compute intensive workloads migrate from CPUs to specialized processors. If CPUs and accelerators were to swap places in the data center, as they did in HPC, the accelerator market would approach Nvidia’s estimate, implying a 10x increase over the next five to ten years.

Artificial intelligence is giving rise to fake fingerprints. Here’s why you should be worried

In the new paper, the researchers used neural networks—the foundational software for data training—to create convincing looking digital fingerprints that performed even better than the images used in the earlier study. Not only did the fake fingerprints look real, they contained hidden properties undetectable by the human eye that could confuse some fingerprint scanners.

Julian Togelius, one of the paper’s authors and an NYU associate computer science professor, said the team created the fake fingerprints, dubbed DeepMasterPrints, using a variant of neural network technology called “generative adversarial networks (GANs),” which he said “have taken the AI world by storm for the last two years.”

Daniel Kahneman: Your intuition is wrong, unless these 3 conditions are met

There are three conditions that need to be met in order to trust one’s intuition. The first is that there has to be some regularity in the world that someone can pick up and learn. “So, chess players certainly have it. Married people certainly have it,” Kahnemen explained. However, he added, people who pick stocks in the stock market do not have it. “Because, the stock market is not sufficiently regular to support developing that kind of expert intuition,” he explained. The second condition for accurate intuition is “a lot of practice,” according to Kahneman. And the third condition is immediate feedback. Kahneman said that “you have to know almost immediately whether you got it right or got it wrong.”

The 80-hour workweek

Every year on Wall Street, first-year analysts who were once courted by college recruiters with lines about how they would be given meaningful, creative work and the chance to learn from top executives are shown the harsh truth: they are Excel grunts whose work is often meaningless not just in the cosmic sense, but in the sense of being seen by nobody and utilized for no productive purpose. Some of the hundred-page pitch books analysts spend their late-night hours fact-checking in painstaking detail are simply thrown away after being given a quick skim by a client. In other cases, the client doesn’t read the deliverables at all, and the analysts’ work is literally garbage.

Most days that winter, he had worked the “banker nine-to-five”—getting to work at 9:00 a.m. and staying until 5:00 a.m. the next morning, at which point he’d trudge back to his Murray Hill apartment, sleep for three or four hours, and do it all over again. Even the money he was making wasn’t cheering him up. Ricardo had once made the mistake of calculating what his hourly wage would be, given his salary and a conservative estimate of his year-end bonus. The result—which he estimated was something like $16 an hour, after taxes—was much more than minimum wage, but not nearly enough for Ricardo to be able to justify the punishment he’d been taking.

The question of in office hours versus out of office hours is a good one in thinking about a career in the finance industry. I knew basically nothing when I came into finance so a massive amount of my learning was done on my own after hours.

Curated Insights 2018.11.02

Steve Jobs had an incredible definition of what a company should be

The company is one of the most amazing inventions of humans, this abstract construct that’s incredibly powerful. Even so, for me, it’s about the products. It’s about working together with really fun, smart, creative people and making wonderful things. It’s not about the money. What a company is, then, is a group of people who can make more than just the next big thing. It’s a talent, it’s a capability, it’s a culture, it’s a point of view, and it’s a way of working together to make the next thing, and the next one, and the next one.


Fossil fuels will save the world (really)

That fossil fuels are finite is a red herring. The Atlantic Ocean is finite, but that does not mean that you risk bumping into France if you row out of a harbor in Maine. The buffalo of the American West were infinite, in the sense that they could breed, yet they came close to extinction. It is an ironic truth that no nonrenewable resource has ever run dry, while renewable resources—whales, cod, forests, passenger pigeons—have frequently done so.


Gundlach: People want to be told what to think. I don’t

My biggest lesson that I’ve learned… I have the same flaw that every human being has and that is: As you’re growing up and getting older, you believe that everybody’s like you. You just extrapolate your personality traits and proclivities on other people. Then you start to realize increasingly, that that’s not true. And I believed, therefore, that everybody was intellectually objective and honest and wanted to figure things out for themselves. And I didn’t understand, for probably as long as 20 years, why I couldn’t convince people of almost mathematically analytical arguments regarding markets. And it was finally after years of this that I realized that people actually want to be told what to think.

It took me a long time to understand that. Not me, see, I don’t want to be told what to think. And so I figured nobody wants to be told what to think. But indeed, I think almost everybody wants to be told what to think. That creates a tremendous advantage in managing money. Because in that window of time between a fact and people being told what the fact means, you have a window if you’re capable of figuring out what it means – and don’t need to be told what it means – where you can actually act before other people and I found I’ve made a lot of money that way.

I remember when Ben Bernanke announced the Fed funds rate was going to stay at 0% for three years, and the markets didn’t move. And I had my traders look for this asset class in the bond market that would be the primary beneficiary of rate staying at zero for three years. And I said, “How much of the prices up?” And they said, “They’re not up at all.”

Assessing IBM’s $34 billion Red Hat acquisition

Dan Scholnick, general partner at Trinity Ventures, whose investments have included New Relic and Docker, was not terribly impressed with the deal, believing it smacked of desperation on IBM’s part. “IBM is a declining business that somehow needs to become relevant in the cloud era. Red Hat is not the answer. Red Hat’s business centers around an operating system, which is a layer of the technology stack that has been completely commoditized by cloud. (If you use AWS, you can get Amazon’s OS for free, so why would you pay Red Hat?) Red Hat has NO story for cloud,” he claimed in a statement.

Forrester analyst Dave Bartoletti sees the cloud native piece as being key here. “The combined company has a leading Kubernetes and container-based cloud-native development platform, and a much broader open source middleware and developer tools portfolio than either company separately. While any acquisition of this size will take time to play out, the combined company will be sure to reshape the open source and cloud platforms market for years to come,” he said.


IBM’s old playbook

The best thing going for this strategy is its pragmatism: IBM gave up its potential to compete in the public cloud a decade ago, faked it for the last five years, and now is finally admitting its best option is to build on top of everyone else’s clouds. That, though, gets at the strategy’s weakness: it seems more attuned to IBM’s needs than potential customers. After all, if an enterprise is concerned about lock-in, is IBM really a better option? And if the answer is that “Red Hat is open”, at what point do increasingly sophisticated businesses build it themselves?

The problem for IBM is that they are not building solutions for clueless IT departments bewildered by a dizzying array of open technologies: instead they are building on top of three cloud providers, one of which (Microsoft) is specializing in precisely the sort of hybrid solutions that IBM is targeting. The difference is that because Microsoft has actually spent the money on infrastructure their ability to extract money from the value chain is correspondingly higher; IBM has to pay rent:

The threat of Amazon’s content strategy

Even if content is created by a publisher and merely distributed through the tech platform, the tech company still captures its data; Netflix, for example, doesn’t share ratings data with TV producers, and Amazon doesn’t share Kindle readership data with the publishing industry. Meanwhile, Facebook actually shared false data with brands about their video’s viewership for years.

  • Anheuser-Busch InBev acquired a stake in RateBeer, a leading beer review platform, and October, a beer culture website.
  • Popular makeup startup Glossier initially launched as a content site; it then used insights gathered from users to develop its own line of cosmetics. Now, it aims to launch a new social commerce platform to encourage user reviews and feedback.
  • L’Oreal invested in Beautycon Media, which creates digital beauty content and hosts festivals for influencers
  • Mattress startup Casper even launched its own magazine; the current issue includes features like “A skeptic’s guide to crystals” and an adult coloring book.

Social Capital’s Chamath Palihapitiya says ‘we need to return to the roots of venture investing’

“The dynamics we’ve entered is, in many ways, creating a dangerous, high stakes Ponzi scheme. Highly marked up valuations, which should be a cost for VCs, have in fact become their key revenue driver. It lets them raise new funds and keep drawing fees.”

“VCs bid up and mark up each other’s portfolio company valuations today, justifying high prices by pointing to today’s user growth and tomorrow’s network effects. Those companies then go spend that money on even more user growth, often in zero-sum competition with one another. Today’s limited partners are fine with the exercise in the short run, as it gives them the markups and projected returns that they need to keep their own bosses happy.”

“Ultimately, the bill gets handed to current and future LPs (many years down the road), and startup employees (who lack the means to do anything about the problem other than leave for a new company, and acquire a ‘portfolio’ of options.)”

The coming storm for consumer staples dividends

AB InBev argued that by taking its leverage down to 2x net debt/EBITDA, it will reduce its cost of capital and “maximize total enterprise value.” All else equal, a lower cost of debt would in theory increase enterprise value, yet AB InBev already has solidly investment-grade credit ratings (e.g., A- from S&P). A ratings upgrade within the investment-grade space would likely only have a marginal impact on lowering cost of debt. Deleveraging could even increase its cost of capital, as more expensive equity takes a greater share of the capital structure.

Ultimately, a company’s dividend should be affordable, reflect the growth in shareholder value creation, and help management more prudently select high-return projects rather than pursue wasteful “empire building” deals. Dividends can be a problem, however, when they become too generous and handcuff management’s ability to invest in high-return projects and defend or widen the firm’s economic moat. When this happens, a dividend “rebasing” or “cut” would benefit long-term shareholders.


Uber-inequality

Uber received proposals from investment banks that pegged the ride-hailing firm’s IPO valuation at $120B. So, that posits Uber’s value is greater than the value of the US airline industry or the US auto industry (excluding Tesla). I love Uber and think the firm is genius. But that valuation is insane. Uber’s model doesn’t have the moats of an auto firm or even Airbnb, which must create global demand and supply (a local competitor to Airbnb doesn’t work, as visitors from other countries wouldn’t know about it). In contrast, local on-demand taxi services abound, even if without an app. The 120K readers of this newsletter could each put in $250, and boom — we have the number-three ride-hailing firm in Miami. Who’s with me?

In today’s economy, innovation means elegant theft: robbery of your data, privacy, health insurance, or minimum-wage protection. Uber has 16K employees and 3M driver partners. “Driver partner” means some great things. It means you don’t have to show up to an office. And it means you can work whenever you want — this is key. When I speak to Uber drivers, I always ask, “Do you like working for Uber?” The overwhelming majority say yes and reference the flexibility. I’ve been especially struck by how many need the flexibility, as they’re taking care of someone who’s sick. So many people taking care of others. So many people loving other people. And it comes at a huge cost. Many of them used to have jobs with benefits. Many had to move to a strange place to take care of their sister, mother, nephew.

The economic value of artificial intelligence

In the near term, around $6.6 trillion of the expected GDP growth will come from productivity gains, such as the continued automation of routine tasks. Over time, increased consumer demand for AI-enhanced offerings will overtake productivity gains and result in an additional $9.1 trillion of GDP growth by 2030.

China is expected to see the greatest economic gains from AI, a $7 trillion or 26% boost in GDP growth. One reason is the high proportion of China’s GDP that is based on manufacturing, where AI is expected to have a particularly big impact between now and 2030. Even more important over the longer term is China’s higher rate of AI investments compared to North America and Europe.

China is expected to see the greatest economic gains from AI, a $7 trillion or 26% boost in GDP growth. One reason is the high proportion of China’s GDP that is based on manufacturing, where AI is expected to have a particularly big impact between now and 2030. Even more important over the longer term is China’s higher rate of AI investments compared to North America and Europe.

In North America, the economic gains from AI are expected to reach $3.7 trillion or 14.5% of GDP growth by 2030. North America will see the fastest growth in the near term, given its current lead in AI technologies, applications, and market readiness. But China will likely begin to catch up by the middle 2020s given its accelerating AI investments.


A.I. is helping scientists predict when and where the next big earthquake will be

Some of the world’s most destructive earthquakes — China in 2008, Haiti in 2010 and Japan in 2011, among them — occurred in areas that seismic hazard maps had deemed relatively safe. The last large earthquake to strike Los Angeles, Northridge in 1994, occurred on a fault that did not appear on seismic maps.

Curated Insights 2018.09.21

Brent Beshore: Learning to pole vault

Marketing will only get you where you’re going faster. If your product isn’t valuable, marketing will help put you out of business, fast. The best way to build trust and generate attention is to be relatively excellent. I say “relatively” because some markets are more efficient/mature than others. The less developed a market, the less valuable you have to be in absolute terms. You just have to be better than everyone else. I don’t want to try to outcompete smart, well-read, and hard working people. I want to find the lowest bar to jump over and then get good at pole vaulting.

Picking your field is arguably more important to your success than your current skill and future capacity. In some segments of business, everyone makes lots of money and the very best do outrageously well. In other areas, even the very best often declare bankruptcy. It’s a base rate analysis. Assume you’re only going to be mediocre, then explore what business and life look like if that’s true. So choose your field wisely and get good at what you’re doing before trying to make noise.

AI has far-reaching consequences for emerging markets

Without a cost incentive to locate in the developing world, corporations will bring many of these functions back to the countries where they’re based. That will leave emerging economies, unable to grasp the bottom rungs of the development ladder, in a dangerous position: the large pool of young and relatively unskilled workers that once formed their greatest comparative advantage will become a liability – a potentially explosive one.

The result will be an unprecedented concentration of productive capacity and wealth in the hands of the elite AI companies, almost all of which are located in the US and China. Of the US$15.7 trillion in wealth that AI is forecast to generate globally by 2030, a full 70 per cent will accrue to those two countries alone, according to a study by consulting firm PwC.

Spotify will now let indie artists upload their own music

According to a recent report by The NYT, artists working with labels may see much smaller percentages. The report said that Spotify typically pays a record label around 52 percent of the revenue generated by each stream. The label, in turn, then pays the artist a royalty of anywhere from 15% to as high as 50%. If artists are dealing directly with Spotify, they could be making more money.

Labels suggested that they could retaliate against Spotify for overstepping. The NYT had also said. They may do things like withhold licenses Spotify needs for key international expansions, like India, or not agree to new terms after existing contracts expire. They could also offer more exclusives and promos to Spotify’s rivals, like Apple Music, which has surged ahead in the U.S. and is now neck-and-neck here with Spotify for paid subscribers.

A music upload feature also means artists who own their own rights could break out big on Spotify if they catch the attention of playlist editors – something that Spotify now makes it easier for them to do, as well. In addition, having indies upload music directly means Spotify could better compete against Apple Music by attracting more artists and their fans to its platform.


Apple’s neural engine = Pocket machine learning platform

If you have followed many of the posts I’ve written about the challenges facing the broader semiconductor industry, you know that competing with Apple’s silicon team is becoming increasingly difficult. Not just because it is becoming harder for traditional semiconductor companies to spend the kind of R&D budget they need to meaningfully advance their designs but also because most companies don’t have the luxury of designing a chip that only needs to satisfy the needs of Apple’s products. Apple has a luxury as a semiconductor engineering team to develop, tune, and innovate specialized chips that exist solely to bring new experiences to iPhone customers. This is exceptionally difficult to compete with.

However, the area companies can try with cloud software. Good cloud computing companies, like Google, can conceivably keep some pace with Apple as they move more of their processing power to the cloud and off the device. No company will be able to keep up with Apple in client/device side computing but they can if they can utilize the monster computing power in the cloud. This to me is one of the more interesting battles that will come over the next decade. Apple’s client-side computing prowess vs. the cloud computing software prowess of those looking to compete.


Tim Cook reveals in interview that the Chinese consumer is different because they don’t carry the burden of the desktop era

China has not experienced the so-called stage of the desktop Internet, but directly embraced the mobile Internet. Therefore, Chinese consumers do not have the burden of the desktop Internet era. This explains to some extent why China’s mobile payment share is so high. In other countries, the mobile payment process is much slower. In fact, they just have no more attempts.”

Perhaps Apple’s delay in advancing Macs and angering the pro community comes from this deep seated attitude that it’s a “burden” holding back the advancement of their iOS agenda.

The best company you’ve never heard of

With no true competitive threats, wide-moat commercial real estate data provider CoStar Group is a borderline monopoly. The other companies in the space are predominately small startups focused on crowdsourcing data. These companies can’t replicate the intangible assets from the vast cost and effort associated with compiling the data the company offers to its customer base.

Given the importance customers place on the underlying data, CoStar also keeps competitors at bay with a switching cost moat source. It’s just too risky to switch sources. Strong platform effects found throughout CoStar’s product offerings earn the company a network effect moat source, too.

The company continues to increase its coverage and boasts that it covers every building in the country, widening the gap between itself and its fragmented competition. The firm recently established itself as a leading provider of rental data with its acquisitions of Apartment Finder and Apartments.com. CoStar is only 30% penetrated in its target market for apartments, so we see room for growth in this area.

Moreover, CoStar is only 15% penetrated in the broker community and 7% penetrated with institutional investors, two groups we can see the firm going after. As several investments are integrated and benefits are realized, we project CoStar’s economic profit to steadily increase over the next several years, reflecting our positive moat trend rating.

Here’s why Yelp and Grubhub could keep rising

“Grubhub is in the early stages of enabling the shift to online of the still offline dominant restaurant takeout businesses and driving the improved consumer experience that comes with it,” they wrote. About “90% of delivery and pickup orders still come from offline, making the phone book, print out menus and walk-ins the number one competitor to Grubhub and its peers.”

How early is the shift? “We estimate Grubhub has about 40% market share of the third-party online delivery/pickup industry which itself we estimate has a 4% penetration of the $250 billion restaurant takeout industry,” they wrote. “Its early mover and scale advantage—about 85,000 restaurants on its platform in 1,600 cities—has allowed Grubhub to offer, in our view, the best consumer value across its competitors.”


Why Yelp could rise 200%

If we can introduce ourselves to those advertisers with a good ‘til canceled $300, $400 a month, $10, $20 a day kind of service proposition, what we’re finding is it really opens up our sales funnel. It makes our product more competitive in the marketplace. It allows us to get into third-party sales channels that we haven’t been in before. And we’re now kind of one quarter into it and we had this quarter, the first quarter, about 140% as many new or net customer additions in this quarter as we’ve had in any prior quarter and kind of 2x the run rate that we’ve normally seen when we were selling the term contract. And, now, we move to the non-term contract.

In the long-term, our tests and our analysis all show that the LTV of a cohort of advertisers that we bring in today will be quite a bit higher. And what we’ve seen in our tests is that we continue to sell the sort of long and strong loyal long-term advertisers under the new pricing model just as we always have, but on top of that we’re introducing ourselves to a lot more new customers along the way

Yelp is in the early days of elevating the consumer experience by expanding the number of transactional features such as Request-A-Quote from a home service professional or book a restaurant reservation or spa appointment. Request-A-Quote lead volume grew 27% from the first to the second quarter of 2018 and topped 5.5 million delivered requests in the second quarter. During that same short timeframe, revenue attributable to Request-A-Quote increased by more than 50%, surpassing a $35 million annual run rate at the end of the second quarter. The company is not yet fully monetizing Request-A-Quote, which we believe could accelerate free cash growth even further. We like finding misunderstood, yet promising, and free embedded call options within the companies we invest in and hope Request-A-Quote proves to a second material avenue for free cash per share growth.


GGV Capital: Unpacking Xiaomi’s IPO

Instead of paying for users, Xiaomi actually gets paid at least 5% gross margin through hardware to get users…it’s a very different model from almost any other internet services model out there. So if this is sustainable, and to make sure this is sustainable is to have a lot more hardware products out there that the middle class can buy, and use that portfolio of hardware devices to get paid for acquiring users, so that internet services can scale thereafter…There’s definitely elements of Muji and Uniqlo in a different field for Xiaomi, there’s definitely elements of a Costco model of subscription plus very low cost to make sure more products are affordable by the rising consumer class, there’s definitely elements of Amazon in there as a platform to sell many products and being very focused at delivering a superior experience…

If we look at the number of internet users coming online, the next 1.5bn internet users coming online between now and 2030, most of that growth will come from the 74 countries that Xiaomi is in already. So when people ask me if Xiaomi is coming to the US or not, they completely miss the point, the growth is coming from the existing countries that Xiaomi’s already in…

Xiaomi has over 18 apps, each with monthly active users of over 50mn. It also has 38 apps, each with over 10mn MAUs. In aggregate, it did over 1.5bn RMB in internet services revenue in 2017, which already puts them as a top 25 internet services only company in the world. The most popular [app] that people know is probably Xiaomi Video, which has an interesting way of becoming aggregation services. It doesn’t license content from anyone, what it does is it aggregates content from all the top Chinese video apps, each of which have already licensed the content and whenever a user clicks on a video, it takes you to the content from its partners but within the app itself, so you can have a more integrated experience. It charges advertising revenue and also subscription from the users…and they share that revenue with its partners that provide the original video content. So, it can focus on providing the most comprehensive collection of content to the user, at the same time, so far, they don’t have to spend much money on acquiring the content itself.”


Tesla, software and disruption

It’s pretty clear that electric disrupts the internal combustion engine, and everything associated with it. It’s not just that you replace the internal combustion engine with electric motors and the fuel tank with batteries – rather, you remove the whole drive train and replace it with sometime with 5 to 10 times fewer moving or breakable parts. You rip the spine out of the car. This is very disruptive to anyone in the engine business – it disrupts machine tools, and many of the suppliers of these components to the OEMs. A lot of the supplier base will change.

We will go from complex cars with simple software to simple cars with complex software. Instead of many stand-alone embedded systems each doing one thing, we’ll have cheap dumb sensors and actuators controlled by software on a single central control board, running some sort of operating system, with many different threads (there are a few candidates). This is partly driven by electric, but becomes essential for autonomy.

Tesla’s first bet is that it will solve the vision-only problem before the other sensors get small and cheap, and that it will solve all the rest of the autonomy problems by then as well. This is strongly counter-consensus. It hopes to do it the harder way before anyone else does it the easier way. That is, it’s entirely possible that Waymo, or someone else, gets autonomy to work in 202x with a $1000 or $2000 LIDAR and vision sensor suite and Tesla still doesn’t have it working with vision alone.

‘Flash Boys’ exchange IEX wins first listing

The U.S. corporate-listings business, in which companies pay fees to an exchange for services tied to being the primary venue for the company’s stock trading, has for years been an effective duopoly of the NYSE and Nasdaq. A third big exchange group, Cboe Global Markets Inc., lists exchange-traded funds and its own shares, but hasn’t made a bid to attract other companies. NYSE parent Intercontinental Exchange Inc. and Nasdaq earned a combined $684 million from listings last year, according to the two exchange groups.

“We at Interactive Brokers understand that being the first listing on a new exchange may entail certain risk, but we think that individual and institutional customers who own and trade our stock will receive better execution prices and that advantage will outweigh the risk,” Mr. Peterffy said in a press release announcing the move.

Because of China’s outsized workforce, the density of automation usage lags other countries: 68 robots per 10,000 industrial workers, compared with 631 bots for every 10,000 manufacturing staff in South Korea, the global leader in automation. Singapore, Germany and Japan all have higher densities of automation than China. China wants to more than double that usage density to 150 for every 10,000 workers by 2020. To do so would require massive amounts of government help.

‘Made In China 2025’: a peek at the robot revolution under way in the hub of the ‘world’s factory’

A skilled factory worker earns about 36,000 yuan a year in wages and benefits in China’s poorer provinces and second-tier cities, away from the coast. Total remuneration can exceed 60,000 yuan in cities nearer the coast and along the eastern seaboard, like in the Pearl River and Yangtze River deltas. A 200,000 yuan robot that can do the job of three humans can recoup its capital cost in 22 months in central provinces, or in a little over a year in coastal cities. In the face of rising prices pressures for labour, energy and rents, such a cost advantage would be attractive to many manufacturers.

China’s total spending on research and development is estimated to have risen 14 per cent last year to 1.76 trillion yuan, according to the Ministry of Science and Technology.

“Among the thousands of so-called Chinese robotics companies – including robot and automated equipment producers as well as those who only provide automation integration solutions – only about 100 firms could mass produce the main body and core components of high-end and middle-market industrial robots, such as servo motors, robot controllers and speed reducers,” he said. “We lack original research and have already tried to catch up by copying advanced technology. But neither technology-related mergers and acquisitions nor copycat [production] can close the gap in the short term.”

He said many domestic robotics manufacturers were still developing the traditional core parts of robots, like servo motors, robot controllers and speed reducers. But these parts would not be the core components of the future, he said.

Don’t take asset allocation advice from billionaires

One of the best ways to stay out of trouble with your finances is to focus all of your energy on your own circumstances and ignore what other people say or do with their money. Not only will it likely save you from making a grievous financial error but it will also make you happier. Constantly comparing yourself or your portfolio to others can be exhausting.

This is how to raise emotionally intelligent kids: 5 secrets from research

Don’t argue the facts. Feelings aren’t logical. You wouldn’t expect the new employee to know how to find the bathroom and you shouldn’t expect a child to know how to handle emotions that, frankly, you still have problems dealing with after decades of experience. Don’t immediately try to fix things. You need to establish you’re a safe ally before you can solve anything. Understanding must precede advice, and, just as with adults, they decide when you understand.

The critical distinction Gottman realized is that it’s important to accept all feelings — but not all behavior. If you skip immediately to problem-solving, the kid never learns the skill of how to deal with those uncomfortable emotions. You want to use “empathetic listening.” Get them to talk. Help them clarify. Validate their feelings (but, again, not necessarily their behavior). They need to feel you really understand and are on their side.

Providing words in this way can help children transform an amorphous, scary, uncomfortable feeling into something definable, something that has boundaries and is a normal part of everyday life. Anger, sadness, and fear become experiences everybody has and everybody can handle. Labeling emotions goes hand in hand with empathy. A parent sees his child in tears and says, “You feel very sad, don’t you?” Now, not only is the child understood, he has a word to describe this intense feeling. Studies indicate that the act of labeling emotions can have a soothing effect on the nervous system, helping children to recover more quickly from upsetting incidents.

As we have discussed earlier, the implications of teaching a child to self-soothe are enormous. Kids who can calm themselves from an early age show several signs of emotional intelligence: They are more likely to concentrate better, have better peer relationships, higher academic achievement, and good health. My advice to parents, then, is to help your kids find words to describe what they are feeling. This doesn’t mean telling kids how they ought to feel. It simply means helping them develop a vocabulary with which to express their emotions.

In an ideal world, we’d always have time to sit and talk with our kids as feelings come up. But for most parents, that’s not always an option. It’s important, therefore, to designate a time—preferably at the same period each day—when you can talk to your child without time pressures or interruptions.

Curated Insights 2018.09.14

Risk, uncertainty and ignorance in investing and business – Lessons from Richard Zeckhauser

People feel that 50% is magical and they don’t like to do things where they don’t have 50% odds. I know that is not a good idea, so I am willing to make some bets where you say it is 20% likely to work but you get a big pay-off if it works, and only has a small cost if it does not. I will take that gamble. Most successful investments in new companies are where the odds are against you but, if you succeed, you will succeed in a big way.” “David Ricardo made a fortune buying bonds from the British government four days in advance of the Battle of Waterloo. He was not a military analyst, and even if he were, he had no basis to compute the odds of Napoleon’s defeat or victory, or hard-to-identify ambiguous outcomes. Thus, he was investing in the unknown and the unknowable. Still, he knew that competition was thin, that the seller was eager, and that his windfall pounds should Napoleon lose would be worth much more than the pounds he’d lose should Napoleon win. Ricardo knew a good bet when he saw it.

…in any probabilistic exercise: the frequency of correctness does not matter; it is the magnitude of correctness that matters…. even though Ruth struck out a lot, he was one of baseball’s greatest hitters…. Internalizing this lesson, on the other hand, is difficult because it runs against human nature in a very fundamental way… The Babe Ruth effect is hard to internalize because people are generally predisposed to avoid losses. …What is interesting and perhaps surprising is that the great funds lose money more often than good funds do. The best VCs funds truly do exemplify the Babe Ruth effect: they swing hard, and either hit big or miss big. You can’t have grand slams without a lot of strikeouts.

Risk, which is a situation where probabilities are well defined, is much less important than uncertainty. Casinos, which rely on dice, cards and mechanical devices, and insurance companies, blessed with vast stockpiles of data, have good reason to think about risk. But most of us have to worry about risk only if we are foolish enough to dally at those casinos or to buy lottery cards….” “Uncertainty, not risk, is the difficulty regularly before us. That is, we can identify the states of the world, but not their probabilities.” “We should now understand that many phenomena that were often defined as involving risk – notably those in the financial sphere before 2008 – actually involve uncertainty.” “Ignorance arises in a situation where some potential states of the world cannot be identified. Ignorance is an important phenomenon, I would argue, ranking alongside uncertainty and above risk. Ignorance achieves its importance, not only by being widespread, but also by involving outcomes of great consequence.” “There is no way that one can sensibly assign probabilities to the unknown states of the world. Just as traditional finance theory hits the wall when it encounters uncertainty, modern decision theory hits the wall when addressing the world of ignorance.


Hank Paulson says the financial crisis could have been ‘much worse’

While Bear Stearns’ failure in normal markets would not hurt the U.S. economy, we believed that the system was too fragile and fear-driven to take a Bear Stearns bankruptcy. To those who argue that Bear Stearns created moral hazard and contributed to the Lehman failure, I believe just the opposite—that it allowed us to dodge a bullet and avoid a devastating chain reaction.

If Bear had failed, the hedge funds would have turned on Lehman with a vengeance. Lehman would have failed almost immediately and the result would have been much worse than Lehman’s September failure, which occurred after we had stabilized Fannie Mae and Freddie Mac and Bank of Americaacquired Merrill Lynch. I would hate to imagine what would have happened if this whole thing started before we’d stabilized Fannie and Freddie.

An interview with Tim Geithner on this topic was done recently at the Yale School of Management and he speaks much more authoritatively on the limits of the Fed powers than I, but here goes. While our responses may have looked inconsistent, Ben, Tim, and I were united in our commitment to prevent the failure of any systemically important financial institution. But we had a balkanized, outdated regulatory system without sufficient oversight or visibility into a large part of the modern financial system and without the necessary emergency powers to inject capital, guarantee liabilities, or wind down a non-banking institution. So we did whatever we could on a case-by-case basis.

For Lehman, we had no buyer and we needed one with the willingness and capacity to guarantee its liabilities. Without one, a permissible Fed loan would not have been sufficient or effective to stop a run. To do that, the Fed would have had to inject capital or guarantee liabilities and they had no power to do so. Now, here’s the point that I think a lot of people miss: In the midst of a panic, market participants make their own judgments and a Fed loan to meet a liquidity shortfall wouldn’t prevent a failure if they believed Lehman wasn’t viable or solvent. And no one believed they were.

AIG is a cautionary tale. We should not have let our financial regulatory system fail to keep up with modern financial markets. No single regulator had oversight visibility or adequate powers to deal with AIG. Its insurance companies were regulated at the state level, its holding company was like a giant hedge fund sitting on top of the insurance companies, and it was regulated by the ineffective Office of Thrift Supervision, which also regulated—get this—Countrywide, WaMu, IndyMac, GE Capital. They all selected their regulator. So you get the picture, it’s regulatory arbitrage.

And I’m concerned that some of the tools we effectively used to stave off disaster have now been eliminated by Congress. These include the ability of Treasury to use its exchange stabilization fund to guarantee the money market funds, the emergency lending authority the Fed used to avoid the failure of Bear and AIG, and the FDIC’s guarantee of bank liabilities on a systemwide basis, which was critical.

The global smartphone supply chain needs an upgrade

At the peak in October 2017, smartphone components accounted for over 33% of exports from Taiwan, 17% of those from Malaysia and 16% from Singapore. Smartphones comprise 6% of Chinese exports. Memory chips flow from South Korea and Vietnam; system chips from Malaysia, Taiwan and elsewhere; and displays from Japan and South Korea. Rich-world firms, such as Qualcomm, sell licences to use their intellectual property (IP). The parts are then assembled, mainly by armies of Chinese workers.

Apple and 13 of its chip suppliers earn over 90% of the total pool of profits from the Apple system. Meanwhile the tail of other firms doing more basic activities must pay for most workers, inventories and fixed assets (see chart). So they have in aggregate a weak return on equity, of 9%, and a net profit margin of just 2%. Their earnings have not risen for five years. They include assemblers such as Taiwan’s Hon Hai and niche component makers, some of which are visibly struggling. On August 22nd AAC Technologies, a specialist in making phones vibrate, said its second-quarter profits fell by 39% compared with the previous year.

Apple, Samsung and most semiconductor makers could ride out such tensions, with their high margins and cash-laden balance-sheets. But the long chain of other suppliers could not, given their razor-thin margins, big working-capital balances and fixed costs. Tariffs could push them into the red. Of the 132 firms, 52% would be loss-making if costs rose by just 5%. And a ZTE-style cessation of trade would be disastrous. If revenues dried up and the 132 firms continued to pay their own suppliers, short-term debts and wages, 28% of them would run out of cash within 100 days.

If you are running a big firm in the smartphone complex, you should be reimagining things in preparation for a less open world. In a decade, on its current trajectory, the industry will be smaller, with suppliers forced to consolidate and to automate production. It may also be organised in national silos, with production, IP, profits and jobs distributed more evenly around the world. Firms will need to adapt—or be swiped away.

The story of Box: A unicorn’s journey to public success

The early days of Box’s selling file sharing and collaboration have largely been replaced by big corporate wins. One measure of Box’s success is its penetration of the Fortune 500—from 52% in the second quarter of 2016 to 69% in the same quarter of fiscal 2019. About 58% of Box’s total revenue comes from enterprises of 2,000 employees or more.

In Box’s recently completed fiscal quarter, it closed 50 deals of more than $100,000, compared with 40 a year ago; 11 deals of more than $500,000, versus eight a year ago; and two deals of more than $1 million, compared with four a year ago. It expects a strong pipeline of seven-figure deals in the back half of this year.

But in encouraging its salespeople to pursue bigger deals, Box increasingly faces competition from deeper-pocketed competitors in a total addressable market pegged at $45 billion, based on market research by Gartner and IDC.

Soccer fans, your team is coming after you

At the time of its 2012 initial public offering, Man United counted 659 million fans worldwide. Analysts estimate the team’s revenue this year will be about 587 million pounds ($763 million) — just $1.16 per supporter. Twitter Inc. has just 338 million active monthly users, yet enjoys revenue of $2.4 billion and a market value of $27 billion.

Digital marketing provides the opportunity for teams to put themselves in the middle of the sale of a service or product. It’s not simply about using a website or an app to sell fans more jerseys or baseball caps. It’s about turning the team into a platform, a way of connecting brands to customers, in the same way as Facebook Inc. and Alphabet Inc. already do.

Much in the way that price-comparison websites charge insurers or credit card companies for connecting them to customers, a sports team could, for example, offer its own exclusive video content with another provider’s mobile phone contract and take a cut of the proceeds. If that meant each fan were to spend just one more dollar a year with the club, it would provide a significant boost to sales.


Alibaba-backed apparel-sharing company YCloset brings sharing economy to a new level

Founded in December 2015, YCloset charges a monthly membership fee of 499 yuan and allows female users to rent unlimited clothes and accessories country-wide. Furthermore, users can choose to buy the apparel if they like to and prices fluctuate according to the rent count. Thus far, 75% of the income comes from membership fees and the remaining comes from sales of clothing. YCloset positions itself as a company that offers affordable luxury, professional and designer brand clothing. The company hopes to have the top famous brand to drive the long-tail brands.

In terms of business model, YCloset gradually shifted from one-time supplier purchase to brand partnerships with clothing companies. Brand partnerships allow revenue sharing between YCloset and their partners. To these clothing companies, YCloset gave them a new revenue, at the same time, they may get consumer insights from the data YCloset collects. In the future, YCloset will have joint marketing campaigns with the brands and assist in incubating new brands.

Autonomous delivery robots could lower the cost of last mile delivery by 20-fold

Last mile delivery – the delivery of goods from distribution hubs to the consumer – is the most expensive leg of logistics because it does not submit to economies of scale. The cost per last mile delivery today is $1.60 via human drivers but could drop precipitously to $0.06 as autonomous delivery robots proliferate.

Autonomous delivery robots are roughly seven times more efficient than electric vehicles on a mile per kilowatt basis. The major costs for autonomous delivery robots are hardware, electricity, and remote operators. Unlike in electric vehicles, the battery is not the largest cost component in slow moving robots. Air resistance is a function of velocity squared, suggesting that a robot traveling at four miles per hour loses much less energy than a car traveling at highway speeds to air resistance. As a result, rolling robots do not require large batteries, lowering both hardware and electricity costs relative to more traditional electric vehicles.

If rolling robots enable last mile delivery for $0.06 per mile, artificial intelligence could be advanced enough to improve their unit economics. A remote operator responsible for controlling robots in difficult or confusing situations probably will oversee roughly 100 robots, accounting for more than half of the cost per mile, as shown below. As autonomous capability improves, remote operators should be able to manage larger fleets of robots, bringing down the costs per robot.


Hospitals are fed up with drug companies, so they’re starting their own

A group of major American hospitals, battered by price spikes on old drugs and long-lasting shortages of critical medicines, has launched a mission-driven, not-for-profit generic drug company, Civica Rx, to take some control over the drug supply. Backed by seven large health systems and three philanthropic groups, the new venture will be led by an industry insider who refuses to draw a salary. The company will focus initially on establishing price transparency and stable supplies for 14 generic drugs used in hospitals, without pressure from shareholders to issue dividends or push a stock price higher.


Harvard Business School professor: Half of American colleges will be bankrupt in 10 to 15 years

There are over 4,000 colleges and universities in the United States, but Harvard Business School professor Clayton Christensen says that half are bound for bankruptcy in the next few decades. Christensen and co-author Henry Eyring analyze the future of traditional universities, and conclude that online education will become a more cost-effective way for students to receive an education, effectively undermining the business models of traditional institutions and running them out of business.

Christensen is not alone in thinking that online educational resources will cause traditional colleges and universities to close. The U.S. Department of Education and Moody’s Investors Service project that in the coming years, closure rates of small colleges and universities will triple, and mergers will double.