Curated Insights 2018.01.28

Amazon Go and the future

In every case a huge amount of fixed costs up front is overwhelmed by the ongoing ability to make money at scale; to put it another way, tech company combine fixed costs with marginal revenue opportunities, such that they make more money on additional customers without any corresponding rise in costs.

To be both horizontal and vertical is incredibly difficult: horizontal companies often betray their economic model by trying to differentiate their vertical offerings; vertical companies lose their differentiation by trying to reach everyone. That, though, gives a hint as to how Amazon is building out its juggernaut: economic models — that is, the constraint on horizontal companies going vertical — can be overcome if the priority is not short-term profit maximization.

Amazon, though, having started with a software-based horizontal model and network-based differentiation, has not only started to build out its vertical stack but has spent massive amounts of money to do so. That spending is painful in the short-term — which is why most software companies avoid it — but it provides a massive moat. That is why, contra most of the analysis I have seen, I don’t think Amazon will license out the Amazon Go technology. Make no mistake, that is exactly what a company like Google would do (and as I expect them to do with Waymo), and for good reason: the best way to get the greatest possible return on software R&D is to spread it as far and wide as possible, which means licensing. The best way to build a moat, though, is to actually put in the effort to dig it, i.e. spend the money.

As for Amazon, the company’s goal to effectively tax all economic activity continues apace. Surely the company is grateful about the attention Facebook is receiving from the public, even as it builds a monopoly with a triple moat. The lines outside Amazon Go, though, are a reminder of exactly why aggregator monopolies are something entirely new: these companies are dominant because people love them. Regulation may be as elusive as Marx’s revolution.

People are using Netflix, Hulu, and Amazon Prime in very different ways

Diet Coke’s moment of panic

A growing consumer focus on health has clearly dented soda’s dominion. Beyond widespread concerns of the dangers of artificial sweeteners, government research has found that daily drinkers of diet soda are at higher risk for strokes and other “vascular events.” While Diet Coke’s new can designs are tall and slender—a possible reference to the body type a diet-beverage drinker seeks—more of them simply don’t trust any kind of soda to be a part of a healthy diet. Between 2000 and 2015, switching from sodas to other beverages saved the country an estimated 64 trillion calories in total—that works out to 71 fewer calories per day, per drinker.

The role of hydration has been outsourced to bottled water and sports drinks, like Gatorade. Getting a jolt of energy has been outsourced to coffee and energy drinks, like 5-Hour Energy. And the satisfaction of a cold liquid fizzing on one’s tongue? That’s been outsourced to the trendy crop of flavored seltzers, like LaCroix.

Nvidia, Western Digital at chips’ frontier

At the same time, Mobley, interestingly, asked if the ISA itself could be an “alternative” to a GPU or a digital signal processor (DSP). O’Connor seemed to indicate that was the case, saying “As they exist today, if you start implementing that kind of functionality — such as vector instructions, for example — you can implement all that functionality using the set of RISC-V extensions, instead of a proprietary instruction set architectures that might have existed up until now.”

That raises an interesting question for Nvidia as it rolls RISC-V out in chips in its next iteration of Falcon. Will an open, shared, standard ISA erode any of the lock-in that Nvidia gets for its GPUs? Or is the “CUDA” programming environment really the important software layer that helps Nvidia maintain and extend its dominance in programming?

Big bets on A.I. open a new frontier for chip start-ups, too

The explosion is akin to the sudden proliferation of PC and hard-drive makers in the 1980s. While these are small companies, and not all will survive, they have the power to fuel a period of rapid technological change.

Nvidia was best known for making graphics processing units, or G.P.U.s, which were designed to help render complex images for games and other software — and it turned out they worked really well for neural networks, too. Nvidia sold $143 million in chips for the massive computer data centers run by companies like Google in the year leading up to that summer — double the year before.

By early 2018, according to a report by Forbes, Cerebras had raised more than $100 million in funding. So had four other firms: Graphcore; another Silicon Valley outfit, Wave Computing; and two Beijing companies, Horizon Robotics and Cambricon, which is backed by the Chinese government.

It is still unclear how well any of these new chips will work. Designing and building a chip takes about 24 months, which means even the first viable hardware relying on them won’t arrive until this year. And the chip start-ups will face competition from Nvidia, Intel, Google and other industry giants.

Sony falls as JPMorgan questions bull case for image sensors

Sony is the global leader in the production of image sensors, camera chips which convert light into digital pictures and videos. Despite a cooling in the smartphone industry, it has benefited from a trend to include multiple image sensors in each phone — a technique used to create better-looking pictures and to power simple augmented-reality functions.

Weak demand for the new iPhone X will hurt Sony, which gets half of its image sensor revenue from Apple, Park wrote. He also said the trend for adopting dual cameras is not as strong as first believed, including among Chinese phone makers, which will further hit Sony’s sales.

In Sony’s latest quarter, image sensors accounted for 9.4 percent of revenue and 22 percent of operating profit.

The biggest electric vehicle company you’ve never heard of

Though it operates in similar sectors as Tesla, the companies are very different strategically. For instance, as Elon Musk’s Boring Company tunnels under cities to address congestion, BYD eyes elevated transportation.

Chinese tariffs and taxes on imported electric vehicles also benefit domestic manufacturers, which capture 93% of the market. BYD has an estimated 30% share. Tesla has 6% share, delivering an estimated 10,000 to 12,000 vehicles to China in 2016. Overall, electric vehicles represent less than 2% of total auto sales in China. Officials, however, aim to phase out fossil-fuel vehicles. BYD chairman Wang Chuanfu was quoted as saying that all vehicles will be electrified by 2030.

Although President Donald Trump has threatened a trade war with China, automobile manufacturing is less susceptible than other industries. Owing to freight rates, manufacturing cars locally within distribution markets makes economic sense. Still, BYD doesn’t currently have plans to sell consumer cars in the U.S. Owing to governmental policies and low fuel prices, Li said the U.S. market isn’t as welcoming to new energy vehicles as China, India, and Europe are.

Electricity from all forms of renewables will be consistently cheaper than fossil fuels by 2020

Today, fossil-fuel power typically costs between $0.05 to $0.17 per kWh. By comparison, consider the global-weighted average cost of electricity generated by various forms of renewables in 2017, as calculated by Irena: hydropower ($0.05 per kWh), onshore wind ($0.06 per kWh), bioenergy and geothermal ($0.07 per kWh), and solar photovoltaics ($0.10 per kWh).

Offshore wind and solar thermal power aren’t yet competitive with fossil fuels, but that should change by 2020, Irena predicts, with the cost of solar thermal falling to $0.06 per kWh and offshore wind to $0.10 per kWh. The drivers will be technology development, competitive bidding systems, and large base of experienced project developers across the world.

Bigger, higher and floating — advances that make wind a better power source

It accounted for close to 40 per cent of Denmark’s electricity mix in 2016 and about 10 per cent across the EU. Wind farms were the leading source of new electricity generating capacity in Europe, the US and Canada in 2015, and the second largest in China.

Despite this, less than 4 per cent of the world’s electricity came from the wind in 2015. That is nowhere near enough to help shift the global economy away from the climate-warming fossil fuels that still supply most of the world’s energy.

The three stumbling blocks to a solar-powered nation

Every hour, our sun bombards the Earth with enough light to satisfy humanity’s energy needs for an entire year.

Cell cost: For solar power to meet 30% of the world’s electricity needs, it will need to fall from its current cost of a dollar per watt of electricity to 25 cents per watt…Perovskite cells can be made from materials that could be radically cheaper than conventional silicon. They can also take on novel forms, such as a tint on windows or thin printable sheets. But they still face significant barriers to commercialization: They tend to rapidly degrade when wet, and scientists can’t create large cells with the same efficiency as the small ones they can make in a lab.

Energy management: It isn’t hard to get to the point where solar is producing too much power at some times of day, and none at all when it’s needed most. The first solar panel added to the grid helps offset midday consumption, but the last one to be added might be completely unnecessary, because the grid might already be saturated when it’s capable of producing the most power.

Soft utility costs: The Energy Department estimates that soft costs contribute as much as 64% of the cost of a solar installation. The rest of the cost is split between mounting hardware for solar panels and the cells themselves.

Why 2017 was the best year in human history

Every day, the number of people around the world living in extreme poverty (less than about $2 a day) goes down by 217,000, according to calculations by Max Roser, an Oxford University economist who runs a website called Our World in Data. Every day, 325,000 more people gain access to electricity. And 300,000 more gain access to clean drinking water.

Curated Insights 2018.01.07

The $100 billion venture capital bomb

Son must deploy $20 billion, or a fifth of the fund, every year for the next five years to meet investors’ terms and their expectations in a market that many already consider overvalued.

Son explained to Hauser that there was a big new wave of computing coming — the sixth, in Hauser’s estimation, following on from the mainframe, the minicomputer, the workstation, the PC, and mobile. This next wave would automate processes in industrial manufacturing and on consumer devices. Son said Arm could uniquely capitalize on this new order as the leading processor manufacturer behind the Internet of things.

“This is the company,” Son said in a televised interview. “No one can live on the earth without chips — it’s in cars, refrigerators, everywhere. So if chips are the things everyone needs, and one company has a 99 percent market share, there must be a barrier. They’re not monetizing well enough. But if I own it, we can monetize it much better. I think the company is going to be more valuable than Google.”

Aside from its 95 percent domination of smartphones, Arm has 34 percent of the global processors market. There are currently 110 billion Arm processors in the world. The company has forecast a total of one trillion by 2035. As the applications get more advanced — be that a car, a washing machine, or a drone — they demand smarter processors, which are more expensive to produce in-house. “We price our fee at a tenth of the cost of what it would cost to develop it yourself,” Thornton notes. “So when you’re staring down the barrel of $100 billion and ten years to develop that processor yourself, we can say it will cost $10 billion from us and you can have it instantly. This is why we have expanded so rapidly over 20 years. One by one, design team by design team, we will become the processor of choice in those markets.”

“Arm Holdings has an insight into the future. When Arm makes a contract with a new business venture, providing the Internet of things for automobiles or farming, Arm will know what is in the pipeline for the Internet of things two years ahead.” SoftBank, in turn, gets a head start on funding companies for a market that doesn’t yet exist.

Analysts say SoftBank, which declined to comment for this article, is at work on vertical integration: Foxconn builds devices, Arm supplies the chips, and SoftBank-owned Sprint and OneWeb, an Internet satellite company, operate the networks on which the devices run. Vision Fund portfolio companies will reap the benefits of these partnerships. SoftBank sits in the middle, introducing high-growth prospects from the fund to one another and to the infrastructure on which their success rides.

Units of time are the new currency

Buffett’s not wrong, but technology has changed the nature of competition. While businesses were once considered only as valuable as the dividends they paid out, the “impenetrable” moats that let companies spit off excess cash are dwindling. A moat today is simply a temporary buffer that helps a company get ahead of the next innovation cycle. When you compound time, you’re creating and recreating value faster than the current innovation cycle.

This is the formula for compounding time into a utility and beyond: 1) Reduce friction for your customers and yourself. Use the time you save to build your utility. 2) Compound time by investing in the ecosystem and getting other companies to integrate with your product. Other companies will integrate with you to save themselves time, building on top of your platform and giving you time to invest in the next great business. 3) Buy other people’s time to defend your utility and stay relevant. Smartly acquiring new products helps you maintain your utility.

Jeff Bezos: “All service interfaces, without exception, must be designed from the ground up to be externalizable. That is to say, the team must plan and design to be able to expose the interface to developers in the outside world. No exceptions.” While this created more work in the short-term, it broke Amazon down into hundreds of micro-services that communicated via APIs. By making all services accessible via API, Amazon drastically reduced the time it took to deploy new features and functionality.

Google’s machine-learning algorithms are reportedly five to seven years ahead of the competition. By keeping TensorFlow to itself, Google would have maintained its lead time — similar to how moats are created by stockpiling assets. But by taking the opposite approach and giving TensorFlow away for free, Google created a utility.

Building a traditional moat will be antithetical to building a great business. The only way to survive is to extract the core of your business and spread it out to compound returns on time. First, you have to save time for your customers and even yourself. Then, you have to invest it forward by co-operating with other products in your ecosystem. Finally, you have to acquire new innovation to maintain your lead.

Why has Waymo taken so long to commercialize autonomous taxis?

To estimate the rate at which passengers will tolerate autonomous taxi errors, we analyzed the manually driven car statistics to set the hurdle. On average human driven cars break down roughly once every 50,000 miles and crash once every 240,000 miles,2 thus offering perspective on acceptable tolerance rates for autonomous vehicle SIFs and UFs.

Supporting this hypothesis, its cars seem to have had difficulty making left turns. One possible explanation is that it has chosen not to vertically-integrate, outsourcing vehicle production to partners like Fiat Chrysler and then taking engineering shortcuts by integrating its sensor suite into a product manufactured away from its controls. In contrast, Tesla’s and Cruise Automation’s (GM) manufacturing operations are vertically-integrated, which could become an important source of competitive advantage.

We are skeptical of that negative conclusion for a number of reasons. Today, Waymo probably is trying to maximize its failure rate to identify faults and root them out. Some stretches of road are trickier and some intersections more difficult to navigate than others.

Getting my fix of Starbucks

SBUX has been successful engendering loyalty from its customers as well- Starbucks Rewards has 13.3mm members in the US and an incredible 36% of all dollars tendered in the stores is transacted through the loyalty program (US Company operated stores).

In the US- the average new SBUX location generates revenue of $1.5mm (average unit volume or AUV) and generates a year 1 store profit margin of 34% or $510k. Based on an average store investment of $700k in the US, this results in an ROI of ~75%. Compare this to a McDonalds with an ROI of ~30%, an average fast casual operator at ~40% or even Chipotle (at its peak before the food illness issues) at ~70%. This means that the average SBUX store earns back its investment a third of the way into its second year – very compelling unit economics. The math likely changes with higher investments in Reserve stores and premium Roasteries in the coming years but if these seek to elevate the overall SBUX experience and thus drive pricing power through the entire system, it’s the right move for the long term.

Starbucks is a well-positioned company led by a smart management team playing “the long game”. While store growth in more mature markets and continuing competition in premium coffee may be a drag to future growth, Starbucks benefits from a moat in the form of a strong brand and a loyal, repeat customer that can be extended into more markets and into more than just coffee. And I believe that this moat is sustainable under the right leadership team that understands that Starbucks delivers an experience that extends far beyond just selling coffee. The sustainability of the moat is predicated on continued investment to elevate the store experience and thus drive pricing power. Management has demonstrated a willingness and enthusiasm to invest and has ample runway to do so while also rewarding shareholders with share repurchases.

How big tech is going after your health care

Now, as consumers, medical centers and insurers increasingly embrace health-tracking apps, tech companies want a bigger share of the more than $3 trillion spent annually on health care in the United States, too. The Apple Heart Study reflects that intensified effort.

Each tech company is taking its own approach, betting that its core business strengths could ultimately improve people’s health — or at least make health care more efficient. Apple, for example, has focused on its consumer products, Microsoft on online storage and analytics services, and Alphabet, Google’s parent company, on data.

Last year, Facebook made it more appealing for pharmaceutical companies to advertise their medicines on the platform by introducing a rolling scroll feature where drug makers can list their drug’s side effects in an ad. Such risk disclosures are required by federal drug marketing rules.

Western Digital, Nvidia on board with ‘RISC-V,’ so pay attention, says Benchmark

Any investor interested in learning how adoption of RISC-V stands to disrupt the CISC and RISC processor domains, including discrete processors and/or processor IP (cores and architectures) embedded within simple MCUs as well as advanced ASICs. Additionally, RISC-V stands to disrupt R&D development roadmaps for merchant and captive SoC companies. For example, if Western Digital truly intends to adopt RISC-V in storage products, Marvell will need to reconsider usage of Arm cores. This could lower the upfront licensing and royalty costs for Marvell; however, it may require a revamping of Marvell’s storage controller design flow. Processor IP companies such as Arm Holdings, Synopsys, Cadence Design, Imagination Tech and even CEVA, Inc. could see an impact.

China removes 1,400 baby formula products from shelves

The regulations, effective Jan. 1, require factories making formula to register those products with China’s Food and Drug Administration and pass safety inspections. Plants are limited to working with three brands, and those brands can make only three different products each. China’s FDA has approved 940 infant-formula products from 129 factories so far, the agency said. That compares with more than 2,300 formulations available to parents before Jan. 1.

That vaulted Nestle, Danone and Reckitt Benckiser Group Plc into the top spots in the $20 billion market, according to Euromonitor International.

Capturing those families will be crucial. With the relaxation of China’s one-child policy, Reckitt Benckiser anticipates about 20 million babies being born annually, which could trigger an annual growth rate of at least 7 percent in the infant-formula category during the next five years, said Patty O’Hayer, a spokeswoman. The company bought Mead Johnson for $16.6 billion last year, and its Enfa and Enfinitas brands were approved for sale. Asia generated half of the Enfa lineup’s $3.7 billion in sales for 2016.

The Paris-based company wants to deploy technology such as laser printing to make tampering more difficult and QR codes to ensure traceability of a product back to the factory — moves intended to assure Chinese parents concerned about food safety.

Cancer deaths fall to lowest rate in decades

While a number of breakthrough, high-cost drugs have improved the outlook for people with some deadly cancers, the biggest cause of the decrease in deaths is that Americans are smoking less. The report found decreased smoking rates, and improved detection and treatment, have led to sharp declines in the rate of lung, breast, prostate and colorectal cancer deaths.

How blockchain technology is redefining trust

‘Regulators will like that blockchain-based transactions can achieve greater transparency and traceability– an “immutable audit trail”,’ Masters says. In other words, it could help eliminate the kinds of fraud that come from cooking the books.

How do typical loans work? A bank assesses the credit score of an individual or business and decides whether to lend money. The blockchain could become the source to check the creditworthiness of any potential borrower, thereby facilitating more and more peer‑​to‑​peer financing.

Consider traditional accounting, a multi-billion industry largely dominated by the ‘big four’ audit firms, Deloitte, KPMG, Ernst & Young, and PwC. The digital distributed ledger could transparently report the financial transactions of an organization in real time, reducing the need for traditional accounting practices. And that is why most major players in the financial industry are busy investing significant resources into blockchain solutions. They have to embrace this new paradigm to ensure it works for, not against, them.

In the patent, Goldman describes SETLcoin as having the potential to guarantee ”nearly instantaneous execution and settlement“ for trades. It would mean all the capital the bank is required to keep in reserve, to hedge against the risk of transactions if they don’t settle, would be freed up.

The blockchain raises a key human question: How much should we pay to trust one another? In the past year, I’ve paid my bank interest and fees, some hidden, to verify accounts and balances so that I could make payments to strangers. I’ve spent thousands of dollars on lawyers to draw up contracts because I am not quite sure how another person will behave (and to sort out a few incidents where trust broke down). I’ve paid my insurance company to oversee the risk around my health, car, home, and even life. I’ve paid an accountant to reconcile an auditing issue. I’ve paid an estate agent tens of thousands of dollars essentially to stand between me, the prospective buyer, and the current owner to buy a house. It would seem we pay a lot for people to lord over our lives and double-check what’s happening. All these ‘trusted intermediaries’ are part of the world of institutional trust that is now being deeply questioned.

Today, it is circa 1993 for blockchain technologies. Even though most people barely know what the blockchain is, a decade or so from now it will be like the internet: We’ll wonder how society ever functioned without it. The internet transformed how we share information and connect; the blockchain will transform how we exchange value and whom we trust.


It’s not the decentralized aspect of Bitcoin, it’s how Bitcoin is decentralized: a cryptographer would have difficulty coming up with Bitcoin because the mechanism is so ugly and there are so many elegant features he wants in it. A cryptographer’s taste is for cryptosystems optimized for efficiency and theorems; it is not for systems optimized for virulence, for their sociological appeal. Centralized systems are natural solutions because they are easy, like the integers are easy; but like the integers are but a vanishingly small subset of the reals, so too are centralized systems a tiny subset of decentralized ones. It may be that Bitcoin’s greatest virtue is not its deflation, nor its microtransactions, but its viral distributed nature; it can wait for its opportunity. If you sit by the bank of the river long enough, you can watch the bodies of your enemies float by.

Gyms ditch machines to make space for free weights

In recent years the 420-location chain has scaled back cardio and weight machines to 50% of floor space from about 66%. The gym devotes the other half of floor space to free weights and functional training, which includes things like kettlebell swings and body-weight exercises with TRX suspension straps. It has also expanded its studio group-exercise classes.

“I prefer to do classes, because the teacher pushes me farther than I would push myself,” she says. “I get bored on cardio machines or on the weight machines.”

The shift away from machines is even more pronounced overseas. In 54 gyms of varying price levels in the U.K., members’ time spent on cardio machines dropped 7% between 2013 and this year, even as the total number of gym visits increased, according to an analysis from Edinburgh-based tracking firm GYMetrix.

The Remarkable Early Years of Warren Buffett (Part 1)

The risks of buying a home that’s too big

“The biggest house isn’t necessarily the best house or even the best investment. An older, smaller home with a shorter commute, bigger lot or greater remodel potential may appreciate more. In fact, many fancy new homes can lose value quickly if a developer builds newer homes nearby, while older areas may have more enduring land value.”

“You want enough space to live comfortably, but you don’t want to heat, clean and pay taxes on space you aren’t utilizing.”

Long-term returns. The money saved in buying a right-size home could pay dividends in the future—literally. A $20,000 savings each year over the life of a 30-year mortgage could result in a nearly $1.2 million nest egg if invested in a stock market portfolio earning 4% a year, Ms. Adam says. The annual savings, when compounded over time, is likely to exceed the appreciation in your home’s value over the term of the mortgage.