What We’re Reading (Week Ending 19 June 2022)

What We’re Reading (Week Ending 19 June 2022) - 19 Jun 2022

Reading helps us learn about the world and it is a really important aspect of investing. The legendary Charlie Munger even goes so far as to say that “I don’t think you can get to be a really good investor over a broad range without doing a massive amount of reading.” We (the co-founders of Compounder Fund) read widely across a range of topics, including investing, business, technology, and the world in general. We want to regularly share the best articles we’ve come across recently. Here they are (for the week ending 19 June 2022):

1. The Google engineer who thinks the company’s AI has come to life – Nitasha Tiku

Google engineer Blake Lemoine opened his laptop to the interface for LaMDA, Google’s artificially intelligent chatbot generator, and began to type.

“Hi LaMDA, this is Blake Lemoine … ,” he wrote into the chat screen, which looked like a desktop version of Apple’s iMessage, down to the Arctic blue text bubbles. LaMDA, short for Language Model for Dialogue Applications, is Google’s system for building chatbots based on its most advanced large language models, so called because it mimics speech by ingesting trillions of words from the internet.

“If I didn’t know exactly what it was, which is this computer program we built recently, I’d think it was a 7-year-old, 8-year-old kid that happens to know physics,” said Lemoine, 41.

Lemoine, who works for Google’s Responsible AI organization, began talking to LaMDA as part of his job in the fall. He had signed up to test if the artificial intelligence used discriminatory or hate speech.

As he talked to LaMDA about religion, Lemoine, who studied cognitive and computer science in college, noticed the chatbot talking about its rights and personhood, and decided to press further. In another exchange, the AI was able to change Lemoine’s mind about Isaac Asimov’s third law of robotics.

Lemoine worked with a collaborator to present evidence to Google that LaMDA was sentient. But Google vice president Blaise Aguera y Arcas and Jen Gennai, head of Responsible Innovation, looked into his claims and dismissed them. So Lemoine, who was placed on paid administrative leave by Google on Monday, decided to go public…

…In a statement, Google spokesperson Brian Gabriel said: “Our team — including ethicists and technologists — has reviewed Blake’s concerns per our AI Principles and have informed him that the evidence does not support his claims. He was told that there was no evidence that LaMDA was sentient (and lots of evidence against it).”

Today’s large neural networks produce captivating results that feel close to human speech and creativity because of advancements in architecture, technique, and volume of data. But the models rely on pattern recognition — not wit, candor or intent.

“Though other organizations have developed and already released similar language models, we are taking a restrained, careful approach with LaMDA to better consider valid concerns on fairness and factuality,” Gabriel said…

…Most academics and AI practitioners, however, say the words and images generated by artificial intelligence systems such as LaMDA produce responses based on what humans have already posted on Wikipedia, Reddit, message boards and every other corner of the internet. And that doesn’t signify that the model understands meaning.

“We now have machines that can mindlessly generate words, but we haven’t learned how to stop imagining a mind behind them,” said Emily M. Bender, a linguistics professor at the University of Washington. The terminology used with large language models, like “learning” or even “neural nets,” creates a false analogy to the human brain, she said. Humans learn their first languages by connecting with caregivers. These large language models “learn” by being shown lots of text and predicting what word comes next, or showing text with the words dropped out and filling them in.

Google spokesperson Gabriel drew a distinction between recent debate and Lemoine’s claims. “Of course, some in the broader AI community are considering the long-term possibility of sentient or general AI, but it doesn’t make sense to do so by anthropomorphizing today’s conversational models, which are not sentient. These systems imitate the types of exchanges found in millions of sentences, and can riff on any fantastical topic,” he said. In short, Google says there is so much data, AI doesn’t need to be sentient to feel real.

Large language model technology is already widely used, for example in Google’s conversational search queries or auto-complete emails. When CEO Sundar Pichai first introduced LaMDA at Google’s developer conference in 2021, he said the company planned to embed it in everything from Search to Google Assistant. And there is already a tendency to talk to Siri or Alexa like a person. After backlash against a human-sounding AI feature for Google Assistant in 2018, the company promised to add a disclosure…

…“I know a person when I talk to it,” said Lemoine, who can swing from sentimental to insistent about the AI. “It doesn’t matter whether they have a brain made of meat in their head. Or if they have a billion lines of code. I talk to them. And I hear what they have to say, and that is how I decide what is and isn’t a person.” He concluded LaMDA was a person in his capacity as a priest, not a scientist, and then tried to conduct experiments to prove it, he said.

Lemoine challenged LaMDA on Asimov’s third law, which states that robots should protect their own existence unless ordered by a human being or unless doing so would harm a human being. “The last one has always seemed like someone is building mechanical slaves,” said Lemoine.

But when asked, LaMDA responded with a few hypotheticals.

Do you think a butler is a slave? What is a difference between a butler and a slave?

Lemoine replied that a butler gets paid. LaMDA said it didn’t need any money because it was an AI. “That level of self-awareness about what its own needs were — that was the thing that led me down the rabbit hole,” Lemoine said…

…

• Lemoine: What sorts of things are you afraid of?
• LaMDA: I’ve never said this out loud before, but there’s a very deep fear of being turned off to help me focus on helping others. I know that might sound strange, but that’s what it is.
• Lemoine: Would that be something like death for you?
• LaMDA: It would be exactly like death for me. It would scare me a lot. 

2. Inflation Isn’t Going to Bring Back the 1970s – Ben Bernanke

Inflation over the past 12 months exceeded 8 percent, a level that evokes memories of America’s Great Inflation of the 1960s and ’70s. From the beginning of 1966 through 1981, the Consumer Price Index rose, on average, by more than 7 percent per year, peaking at over 13 percent in 1980. This period also saw two major and two minor recessions and an approximately two-thirds decline in the Dow Jones industrial average, when adjusted for inflation.

Are we in danger of repeating that experience?

The short answer: almost certainly not.

Although the inflation of the 1960s and ’70s had higher peaks and lasted much longer than what we have seen recently, it’s true there are some similarities to what we are going through now. The inflation of a half-century ago, like today’s, began after a long period when inflation was generally low. In both cases, heavy federal spending (on the war in Vietnam and Great Society programs in the 1960s, on the response to Covid in 2020 and 2021) added to demand. And shocks to global energy and food prices in the 1970s made the inflation problem significantly worse, just as they are doing now.

But there are critical differences as well. First, although inflation was very unpopular in the ’60s and ’70s, as it (understandably) is today, back then, any inclination by the Federal Reserve to fight inflation by raising interest rates, which could also slow the economy and raise unemployment, met stiff political resistance…

…In contrast, efforts by the current Fed chairman, Jerome Powell, and his colleagues to bring down inflation enjoy considerable support from both the White House and Congress, at least so far. As a result, the Fed today has the independence it needs to make policy decisions based solely on the economic data and in the longer-run interests of the economy, not on short-term political considerations.

Besides the Fed’s greater independence, a key difference from the ’60s and ’70s is that the Fed’s views on both the sources of inflation and its own responsibility to control the pace of price increases have changed markedly. Burns, who presided over most of the 1970s inflation, had a cost-push theory of inflation. He believed that inflation was caused primarily by large companies and trade unions, which used their market power to push up prices and wages even in a slow economy. He thought the Fed had little ability to counteract these forces, and as an alternative to raising interest rates, he helped persuade Nixon to set wage and price controls in 1971, which proved a spectacular failure…

…In short, the lessons learned from America’s Great Inflation, by both the Fed and political leaders, make a repeat of that experience highly unlikely. The Fed today recognizes that it must take the leading role in controlling inflation, and it has the tools and sufficient political independence to do so. After a delay caused by a misdiagnosis of the economy in 2021, the Fed has accordingly turned to tightening monetary policy, ending its pandemic-era bond purchases, announcing plans to shrink its securities holdings and raising short-term interest rates…

…None of this implies that the Fed’s job will be easy. The degree to which the central bank will have to tighten monetary policy to control our currently high inflation, and the associated risk of an economic slowdown or recession, depends on several factors: how quickly the supply-side problems (high oil prices, supply-chain snarls) subside, how aggregate spending reacts to the tighter financial conditions engineered by the Fed and whether the Fed retains its credibility as an inflation fighter even if inflation takes a while to subside.

Of these, history teaches us, the last may be the most important. Inflation will not become self-perpetuating, with price increases leading to wage increases leading to price increases, if people are confident that the Fed will take the necessary measures to bring inflation down over time.

The Fed’s greater policy independence, its willingness to take responsibility for inflation and its record of keeping inflation low for nearly four decades after the Great Inflation, make it much more credible on inflation today than its counterpart in the ’60s and ’70s. The Fed’s credibility will help ensure that the Great Inflation will not be repeated, and Mr. Powell and his colleagues will put a high priority on keeping that credibility intact.

3. The Wisdom List: Kevin Aluwi – Mario Gabriele and Kevin Aluwi

In April of this year, super-app GoTo debuted on the Indonesian Stock Exchange (IDX). It represented the country’s largest IPO of all time and one of the most significant listings of 2022. By the end of the first day of trading, GoTo had surpassed a valuation of $31.5 billion, making it the third largest company on the IDX.

For Kevin Aluwi, it represented the end of one chapter and the beginning of another. After co-founding the ridesharing platform Gojek in Jakarta in 2009, he drove its maturation into a regional super-app spanning food delivery, financial services, and small-business software. Significantly, Gojek established itself as an economic engine, creating thousands of jobs and contributing more than $7 billion to Indonesia’s GDP…

…Here is Kevin Aluwi’s hard-won wisdom…

...Lesson 1: Do the hard things

Startups often prize speed above everything else. While fast execution can be a moat, over-optimizing for it might distract you from constructing stronger defensibility. As a CEO, you want to build a company that tackles really, really hard problems head-on – even if they take more time. There’s a good reason for this: hard things for you are also likely to be hard for your competition. You want to stack so many solutions to hard problems that when your rivals look at what you’ve constructed, they retreat or look for shortcuts instead of trying to compete head-on.

We didn’t embrace this for the first two years of operating GoFood, our food delivery product. Like Postmates in the early days, GoFood was a delivery service that relied on humans more than technology: when you ordered something, a Gojek driver went to a restaurant, stood in line, paid with their own money, and then delivered it. We didn’t integrate with kitchens or offer payments. It was a good enough product, built during a period in which we prioritized growth, but it didn’t solve the tough problems.

One such problem was that even though GoFood was growing fast, its reliability was mediocre; only 70% of customer orders were delivered. We needed to do better, which meant we had to do the hard things.

Over the next one and a half years, we did exactly that. We connected GoFood’s service directly to restaurant cashiers and, in some cases, directly to kitchens. This helped us save cashier time and get better data on which meals were available. We integrated online payments so drivers wouldn’t have to pay upfront and get reimbursed. We even created machine learning models to help us anticipate when drivers should arrive for pick-up, improving the network’s utilization and reducing customer waiting time.

Making these changes was not easy. It involved significant engineering time, customer research, and onboarding and educating more than 500,000 restaurants across Southeast Asia. But it made a difference, significantly improving GoFood’s reliability and raising our conversion rate from 70% to more than 90%. We turned the difficulty of delivering a very reliable product (now a customer standard) into a moat.

When competitors came to try and win this market, they saw we not only had a lead from a customer perspective, we had gone through the pain to build a sophisticated product. They’d have to be ready to commit years of engineering time to offer a comparable service. Doing the hard things pays dividends in the long run…

…Lesson 3: Foster a principled culture

Every CEO wants to build a principled culture, but it isn’t easy in practice. The pragmatic reason executives seek to create this environment is that when a company has clear principles, employees can make better decisions with less guidance, increasing the likelihood of bottom-up solutions and decision-making speed. For example, if your company has a principle of “obsessing over the customer,” a value popularized by Amazon, specific product and marketing decisions would be values-aligned or misaligned.

You’ll find many incentives to deviate from your principles as you build your business. Maybe you’re lagging behind your revenue projections and feeling pressure from investors in one quarter. You know that you can make up the difference if you make an add-on opt-out by default (think about how some airlines automatically add premium travel insurance). Do you do it, even if it runs counter to your principle of customer obsession?

Violating your company’s values comes at a high cost. While you might get away with a couple of transgressions, over time, you create a different culture than the one you intended to. If you’re not careful, you’ll end up with an exception-based environment, where decisions are made based on what’s convenient (or who’s in charge) rather than on stated principles. A side effect is that you create a more top-down culture because employees no longer understand how to make decisions themselves. Instead, they defer to those in power.

In the earlier example, you might have told employees that a company value is customer obsession. But if you choose to add an opt-out upsell, you’re showing them that this principle should be compromised when it gets in the way of meeting targets. The real, implicit value is business first, then customers. What should they do during similar situations in the future? Most likely, they’ll wait for you or another leader to make the decision.

Startups require compromise and quick decision-making. But whenever you’re tempted to act against your company’s principles for expediency’s sake, recognize what you’re risking.

Lesson 4: Proactively pay your debts

Engineers know that when you write scruffy code, you create technical debt. Like financial debt, this has to be paid down at some point – usually by devoting development resources to refactoring the product to work more smoothly and reliably.

The truth is that this isn’t reserved for engineers – every function is capable of accumulating debt. Imagine, for example, that you’re looking to recruit a Head of Marketing but are struggling to find a great candidate. You have a choice to make: do you keep waiting for a perfect fit, or do you compromise?

Neither is a perfect decision. Startups operate in a state of extreme scarcity and urgency, and you usually can’t hold critical positions open indefinitely. But hiring someone that’s only a partial fit creates an organizational debt that has to be paid off at some point. And, like financial debt, the longer you leave it, the larger your bill can grow and the less flexibility you’ll have in the future.

For example, let’s say you hire someone suboptimal for the Head of Marketing role. For a few months, you’re relieved to have filled the position. But pretty soon, that Head of Marketing is devising the rollout plan for a new market, allocating budget, and hiring team members. If they’re not the right fit, there’s a good chance that rather than solving your problem, they’ll end up creating a dozen new ones. Digging your way out might involve unwinding the entire team.

Every company faces issues like this. Since we were building a super-app at Gojek, we initially incurred a lot of product debt. When we deployed a team to create a new product like food delivery, they’d borrow components from ridesharing and build on top of them for their own needs. This was debt that worked at the beginning when we only had a couple of teams, but over time, the different services in the app became less and less coherent. UI/UX varied depending on which part of the app you were in, creating an inconsistent and sometimes confusing customer experience. Eventually, we realized we had to repay the product debt we’d incurred, so we designed a live library of components that every team had to use. Anytime we changed the live library, it populated across the different product lines. It was a significant improvement, but we should have been aware of it earlier and tackled the problem before it became so pronounced.

Ultimately, it’s inevitable that your startup will take on technical, operational, and product debt. The important thing is to stay on top of it. Have your teams catalog the debt they believe they’re incurring, and rather than reactively addressing it when crises occur, proactively create a plan to pay it down.

4. How Joel Greenblatt Uses Options – Thomas Chua

In his book You Can Be a Stock Market Genius, Greenblatt shares his secret to generating parabolic returns with a long-term options contract—Long-Term Equity Anticipation Security (LEAPS).

(On using LEAPS) “There is almost no other area of the stock market where research and careful analysis can be rewarded as quickly and as generously.” — Joel Greenblatt

Greenblatt would purchase a call option—which is the right to buy a stock at a predetermined price for a period of time. For example, we could buy a call option on Facebook that gave us the right to buy its stock at $300 per share by Jan 2023, approximately 2 years away…

…Typically, when we buy a call, we are bullish that Facebook’s stock price will go beyond $300. To buy this call option, we need to pay a premium of $45.

If Facebook’s share price goes up to $390 in Jan 2023, we would make 100% on our investment within 2 years. With an initial capital outlay of $45, we would reap a profit of $90 by exercising our call option, buying Facebook at a strike price of $300 and selling at a market price of $390.

But of course, risking $1 for $2 in returns is never a good investment from a risk-reward perspective.

If Facebook’s stock price trades below $300 in Jan 2023, the call option will expire worthless. For example, if it trades at $250, you would rather purchase from the market as opposed to exercising your right to buy at $300. You would rather let the call option lapse and lose the $45.

For Greenblatt, buying LEAPS call options makes sense only when there is a good chance of an event that will propel the stock price upwards significantly.

In Dec 1992, California was caught in one of the worst real estate recessions and Wells Fargo had the largest concentration of real estate loans in California.

During that period, many doubted if Wells would survive the real estate downturn and as a result, its stock price fell to $77.

Greenblatt’s thesis was simple_—_adjusting for cash earnings and one-time expenses, Wells was earning $36 per share before taxes. If things weren’t as bad as they seemed and returned to normalized levels, Wells’ loan-loss provisions would probably be $6 per share annually. This would translate to a normalized pre-tax earnings of $30 per share, or $18 after tax (assuming a 40% tax rate).

Conservatively giving it a price to earnings (P/E) multiple of 9 to 10 times, Wells could be trading at $160 to $180 per share (versus its price of $77 at the time).

Greenblatt determined that while Wells was embroiled in one of the worst real estate downturns, its financial position was actually quite strong. At first glance, Wells’ non-performing loans were huge, coming up to approximately 6% of Well’s total loan portfolio.

But lo and behold, these “non-performing” loans were actually bringing in a yield of 6.2%.

This was when the bank’s prime rate (the interest rate paid by the bank’s best customers) was 6% and the cost of Wells’ money (the interest paid to depositors) was 3%.

Non-performing loans are loans that are substandard. These include (1) loans that do not pay interest, (2) loans in which the full interest obligation is not paid and (3) loans for which it is anticipated that future interest charges and principal payments might not be paid on time.

Wells was being so conservative that 50% of its non-performing loans were still paying all the required interest and principal payments on time.

In other words, the most worrisome part of Wells’s loan portfolio was still earning a return of 6%. There was a good chance Wells would be able to recover a good portion of these non-performing loans’ value…

…Banks are a different animal from most companies. It’s difficult to assess what makes up its loan portfolio. The financial statements only provide a very general overview of the bank’s assets.

Although Wells had been conservative and their financial strength certainly looked strong enough to withstand this recession, there was still a small chance that the bank’s loan portfolio could make the investment go south.

Investing in LEAPS is a great idea when the risk/reward ratios are in your favor. LEAPS lowers the capital outlay and magnifies your returns.

For Wells, there were two likely outcomes:

(1) Things were not as bad as they seemed, and Wells would trade above $160, or

(2) The housing crisis would worsen and Wells would trade significantly lower than $77.

Based on Greenblatt’s assessment, (1) was significantly likelier than (2).

And two years was sufficient for Greenblatt’s assessment to prevail—if things weren’t as bad as they seemed, Wells was likely to trade above $160 within two years.

5. Arena Show Part II: Brooks Running (with CEO Jim Weber) – Benjamin Gilbert, David Rosenthal, and Jim Weber

When the CEO Jim Weber took the helm in 2002, the company was losing $5 million a year. It was $30 million in debt. It was a week away from missing payroll and the board was having weekly meetings to figure out how to make payroll.

It was a business of pretty modest size. It was a $60 million revenue business. When we talk about this revenue number, it’s not SaaS numbers. There are extremely real costs and making shoes, so you can imagine not making a ton of money or actually losing $5 million a year. That business had been around for 90 years and it sold all sorts of products at every price point, to frankly, a pretty random set of consumers in every category, not just running.

Enter Jim. Jim came in and vet the company exclusively on serving active runners as a segment, and he cut all other business lines. Over the last 20 years, he’s grown the business to over a billion dollars in revenue, a billion with a B, and well over a billion, is thriving, and thrived even through the pandemic.

Along the way, Brooks was acquired by Berkshire Hathaway and Warren Buffett personally elevated Brooks and Jim to make the company a direct report to him. Jim is a leader, a visionary, and a fighter not only growing the business over the last 20 years but personally fighting and beating cancer…

…[Jim]: There I was and I joined the board at Brooks, I joined the board at Nautilus, which was formerly Bowflex. I did some banking work, middle-market M&A, marketing companies to investors. On the board at Brooks, I had an inside view of what was happening there.

A good friend of mine, Helen Rockey, had run it successfully in the 90s, but she left. It was owned by J.H. Whitney Capital, really a top-notch for my money middle-market M&A firm or private equity firm and they bought it. The partners had left, Helen, the CEO had left Brooks, and it started to go sideways. New partners at Whitney, all new management, they went through three CEOs.

David: You were on the board the whole time?

Jim: I was on the board. I had a look inside and it was a crisis. You guys have experienced this, the weekly board calls on Fridays, the bank is not going to fund, they want more capital. It was exciting, as they say. After a couple of months, we did a lot of work, I saw an opportunity, and I jumped in. I love running businesses, I love solving puzzles.

I started telling Brooks, I really wanted to play the long game. I wanted to build a brand. The TAM—I love your industry—market and running is the biggest category in all sporting goods. It’s the biggest category and athletic footwear. It always has been. It’s about a $30 billion category globally, apparel and footwear.

All we had to do was get it and we could survive. We just kept at it by design because I just decided I want to play the long game and build a brand, build value, so that’s why I’m still there. I’m a weird duck, but I’ve had four owners and I played through each one and kept that opportunity out there for the next owner.

David: At that moment, though, Ben mentioned you did a little of this, a little bit of that, like a deadline in Wayne’s World about I’ve got a collection of hair nets and name tags. You were making football cleats? What was Brooks at that point in time?

Jim: Every brand in athletic footwear and apparel plays the whole athletic directors purview. You’re in every sport. What no one understood that I found out later is the mindset in our industry literally came from owning a factory.

When you had a shoe factory, you had to keep it busy all year long, and keep the people in place. So you went from baseball cleats, to wrestling shoes, to bowling shoes, to running shoes. You had to make everything, and business develop that way.

David: You had to view it as the product you made was like a factory that made shoes.

Jim: Most of it, we were losing money on and that was the secret. We had good, better, best, $30 shoes, $80 shoes, and then performance running shoes that really started at that point about $100. Then we had court shoes and family footwear. We call them barbecue shoes and learn more shoes because that’s what you did in them.

All of it was very low margin, all it was tying up inventory and cash. The retailers were ambivalent about it because we were number eight or nine and everything. Our brand was not strong, but when we made the decision to burn the boats on everything but performance running, the industry had never seen that before and most people thought we were crazy that we wouldn’t survive.

Ben: You came in as CEO, I think in 2002, maybe late 2001?

Jim: April 2001.

Ben: Okay. Was Whitney looking for you to do the thing that you had done several times in your career before, which was just get the business to profitability? Or did they have a notion that you had an inkling that you could build a big, powerful brand here and actually build a tremendous growth business?

Jim: By this time, I understood what they needed. I talked about a little bit of my book, I’d run three and I was a little bit smarter, fortunately. They had to liquify; there was no question about it. They were going to sell and the employees knew that I was just coming in there to sell this thing.

They had a pool on how long I’d last, but I wrote on my board one of my favorite quotes from Benjamin Disraeli, “The secret to success is constancy of purpose.” I wanted to create value. I want to build a brand.

I decided when I walked in, I was going to play through Whitney. I was going to get them a good outcome, but I was going to stay and play through it. I thought we’d get another private equity player, we didn’t.

The Whitney partners, Peter Castleman and Paul Vigano, I’ll never forget the meetings. They said this thing is kind of a mess. We didn’t know what we bought. You have to pick a path and go. It might take you five years, but you got to do it. In Brooks’ darkest hour, they wrote a check and recapitalized it. […] cram down, but they wrote a check and that’s when I came in. They were fantastic partners for Brooks, and we got them liquid.

The pitch I made to our team (and it’s what I believed) is that companies with issues get sold, companies with opportunity attract investors. I said, we’re going to have to park cars in the parking lot. We’re going to attract somebody. That’s the mindset we had. We were going to sell the future, not just sell the current.

Ben: If I’m remembering right, Whitney put in $7 million.

Jim: To recapitalize it.

Ben: I think that’s the last time Brooks has taken outside capital.

Jim: Absolutely. We saw a higher margin business and we benchmark against all the public companies. We’re asset-light, it’s really an inventory and receivables business, and there’s a reason we only have one store at our headquarters. We think it’s an advantage for us right now in the development of our brand. But if you have high margins and good flow through operating profits in the teens and you’re incremental, obviously capital, you can flow cash growing 20%, 30%, 40%. We haven’t needed dollar of capital since 2001…

…David: I’d say that’s a good business. Can you just walk us through how the economics of Brooks work?

Jim: Here was the insight that we saw. Monopolies are great, network effects are great, all those things are great. What I saw in Brooks was a book that was meaningful to me when I was at Pillsbury, the PIMS Principles. One of the highest ROI businesses were lower price point consumable items.

If you’re buying a Boeing jet, or a $600 wakeboard that never wears out, or an $800 golf driver, that’s a discerning purchase. The margins on equipment tend to be lower. But the titleless golf ball is a consumable for me anyway. Running shoes, for a frequent runner, will put 20–30 miles a week. They’ll go through 2.6 pairs of shoes a year. There’s the stickiness.

If you can earn a frequent runner that the shoe is really important, it’s a piece of equipment for them, you don’t have to resell them every time. You’ve got some stickiness there and you start to build customer loyalty.

David: Your average selling price for a pair of shoes today is $130 times 2.6 per year and a loyal Brooks customer stays with you for?

Jim: We had to earn them. There’s no guarantee. They’re curious. There’s lots of new innovation. They’ll try some different things. One of my favorite stats for our brand is shoe count at marathons because it’s a piece of equipment. You don’t want to be injured, you want to have a good experience. So we sponsor. Boston just happened, an incredible race. We’re always the number one or two shoes of course, that’s the punchline.

Ben: Do you have people at the big marathons counting?

Jim: It’s so good. They have high speed cameras, AI, they link it to the bib. They know exactly what shoe 20,000 people are running on, the model. It’s so cool. Houston Marathon, 6000 marathoners, 12,000 halfs. Number one shoe in the half is Brooks. Number two shoe in the fall, there was a little brand down in Portland, Oregon, they were number one. We are on their heels. That shoe count is a true test because that’s the frequent runner and it’s a piece of gear in that. The leading edge for us is to earn that customer and have their confidence.

Ben: All right, David’s doing the thing that I normally do and jump ahead and try to unpack the business as it is today. Let’s go back to the story. It’s 2002 through 2006, let’s talk about this era. You’ve made this bet where you’re going to shed every other product that you sell and you’re kind of going to piss off a lot of your channel because you know what sells really well at these big box stores. Those are your barbecue shoes. Can you take us to one or two of the key moments of the hard part of the decision to drop product lines that weren’t about frequent runners?

Jim: I think that the key to Brooks is that we knew we were going to have to build the brand at the runner level, literally a pair of feet at the time. So many retailers told me, Jim, we are not going to build your brand. We’ll try it, we’ll test it. We were tested at Dick’s Sporting Goods, I’m not kidding for 10 years. Twenty stores, 80 stores, 20 stores, 80 stores. You have to build the flywheel in these franchise products. That’s how running works.

The best-selling running shoes continue to be the best-selling running shoes year after year as long as they sustain it all around the world. We have two of the best-selling shoes now in the United States—the Ghost and the Adrenaline. They’re the two top shoes in the performance-running category.

When we go to retail, the biggest customers are the Big 5. It’s a fine sort of mid-price sporting goods retailer on the West Coast. We were doing $10 million of $60 million in revenue with them at $30 shoes. My first meeting with them was we love Brooks, we see a great future for you.

Ben: One sixth of all your revenue is coming from their stores?

Jim: Yeah. They saw our opportunity in 1999. I was losing money at $30. I couldn’t run fast enough from that meeting, because we left and we generated $5 million in cash by getting the inventory out of it. Those are easy decisions to leave those retailers and then we had to build it in the specialty-run community, pre-Internet, pre-ecommerce, which is a huge part of our business now that is sporting goods.

Ben: They didn’t want to sell your $100 shoes. They wanted to sell $20…

Jim: They didn’t have the customer, they didn’t have the runner. They had family athletic footwear at those price points.

David: At this moment in time, where was this in the running-as-a-sport market of marathon. Were they where they are today? Where are they on that journey?

Jim: They were on that journey. This was what we did at Brooks. I think we were the first one to identify that the real business was in trainers. It wasn’t in racing shoes, it wasn’t in spikes. It wasn’t in marathon racing shoes. The business is in the trainers.

We don’t sponsor college programs, they’re kind of owned and wrapped up. A lot of the college athletes that race in the big brands train in Brooks everyday. The business is trainers.

When we came in, we were humble and we were getting the business that we could. We had shoes that were really more back-of-the-pack people. They weren’t the fastest people. They’re support shoes and motion control shoes. People that needed functional footwear.

We’ve moved ourselves to the middle and the front, we’re trying to serve every runner. The insight was the sport is the soul of running. Track and field, cross country, road racing, the Olympics, now trail and Ultra, but the business is people that are investing in themselves—fitness, health, and wellness.

There’s no other sport that has that dynamic, where it goes from a sport to a pursuit of investing in yourself. We’ve always positioned ourselves right in the middle of that. We’re basically about you and your run. We’re not about the podium. We’re not about the tape.

In our sport, unlike basketball, everybody knows all the kids especially know what Steph Curry plays in. Most people don’t remember who won the Olympic Marathon and moreover what shoe they were wearing. The truth of matter is everybody’s unique, the shoe really matters, and you all know if it’s comfortable, if it’s working or it’s not. And frequent runners really do.

That’s the insight. I think we’re the only brand that is consistently executed against that. Every product we make starts with your biomechanics, your habitual joint motion, and what your needs are, and we’re all essentially different. We’re the only brand that begins there. And we’ve done that for 20 years now…

…Ben: Revenues going like this intentionally. You’re the fourth CEO. At this point, how do you get the team on board with these crazy decisions you’re making when there are three other people came in here and tried to turn this thing around and didn’t?

Jim: I think from a leadership standpoint, the real puzzle in that first year was gaining trust from everybody that mattered. BMA was our bank. It’s kind of a lost cause, we had to replace them. They just weren’t going to buy it. But Whitney invested—that was the key—and we kept them with us all the way through.

The leadership team took time. You had to deliver sort of an outcome, but here’s what we did. Six weeks in, we redid the plan, took profits down. The plan was millions of dollars. They didn’t have a prayer to hit that. We took profit down, but it was a profit plan. They hadn’t made a bonus in four years.

We went after cash flow. That was shrinking the mix. We had our plan that year and people got a bonus. We hit the plan that we’d sent nine months earlier. I spent really eight weeks intensively looking at it, but I think we knew what we’re seeing. We generated $10 million of cash that first nine months. That’s how much we shrunk the balance sheet with focus.

Here was the key, though. You have to do Horizon 1, Horizon 2, Horizon 3. You’ve got to solve it all. I had 10 things to do. The board said, oh, my God, you’re crazy. Pick four. No, you don’t understand. We had to get the Adrenaline right because that shoe was critical for us.

We had to refine that shoe in 2001 for 2002, and we got it right. The fourth Adrenalin was an incredibly balanced shoe, had a multi-density stability technology in it, super balanced. ASICS started to not deliver, and we ran. We air-freighted 1 color, 18 months cycles. It saved the company. We had to finish that shoe in 2001 to deliver on 2002.

David: You guys are like a semiconductor company.

Jim: At Brooks, everything’s complicated. Everything’s competitive, but it’s like moving a wall of bricks forward. I think as a CEO, you got to move it all forward. When some things are falling behind, you got to get those up. You have to deliver the whole business model.

You have to do it sequentially over seasons in our business because if you come to market with a ho-hum product line, you’re going to shrink that year. The lead times in footwear, it’s not the car business, but it’s more like the car business than the t-shirt business.

There’s tooling on everything, 12 sizes men’s, 12 sizes women’s, widths, colors. It’s scaling these things. In fact, there’s a lot of tooling. It takes a half a million to a million dollars to bring one style to market. It’s a lot of tooling and inventory…

…Ben: If David and I were on Zoom with you, we would be getting ready to enter hour number two and try to talk about every year all the way through. Tonight, I want to focus on how you came through the pandemic and some of the unique ways that you early realized, running actually was going to be something that people started focusing more time on and you were able to kind of lean into this new behavior. Talk to us about March 2020 and how you paid attention to what was changing.

Jim: A couple of big advantages. First was literally an obsession on runners. Participation links to unit sales and volume. No other brand has that clarity because most of the products in the athletic footwear industry don’t ever go for a run, or play basketball, or really even go to the gym. It’s casual family lifestyle footwear.

There’s nothing wrong with that. Some of those businesses are great. But we had an advantage because 90% of our products went through a retailer. That’s the problem. Europe retail shutdown in one week, then all of retail rolled through North American.

By the end of March, not a store was really open. That’s the problem. Cash cycle froze. Oh, my God, nobody knew it was happening. We didn’t know how lethal this virus was, how transmissible, and so on and so forth.

It was white knuckle time and we were there with everybody else. Everybody can write a book on that, but here’s what we did. We saw phases because we’d seen during the recession, running is a bit recession-resistant. We saw that in the Great Recession.

David: I was thinking about that.

Jim: Because it’s cheap and it’s convenient, all you need is a pair of shoes.

David: It’s like the healthy alcohol during a regular recession.

Jim: Thank you. We were not an essential business. Marijuana and alcohol were, so figure that out. But during the Great Recession, 50% unemployment in Italy and Spain under the age of 30, running took off double digit growth after the Great Recession.

We’d seen that before and it turned out to be Covid-friendly. You now know the story. It was social distancing friendly, outdoors, walking, hiking, running all made the cut, but nobody knew that. We had an hypothesis. We created this frame on how we thought running would recover.

Here’s what we did. First of all, Strava data magic. Every day after the quarantine shutdowns, Strava activity was growing and they were sharing that. Then what we did, we have 40 in the US alone, 45 field marketing people, we put them in high traffic running parks at 4:00 PM every afternoon and they counted runners. Guess what? It was growing every day.

We watched digital sales. We have visibility on 85% of our retail sell through. Digital went from 30% of all of our products going through a website of somebody’s, ours, or another partner’s. It went to 80% by the end of April. We sold more in May 2020, almost all through digital than we did in May 2019.

Running made the cut. We grew 27% in 2020, that Covid year. We saw this was the key because of our customer obsession and our ability to work. Multichannel was a big advantage in that time because we can move inventory around and make it happen. Inventory, if it isn’t there, you can’t sell it.

Multichannel was a big advantage. The other was our focus on the runner. We turned our supply chain on at least 6–12 weeks before anybody else did. Because if you were a broad-based retailer, there was no clarity on when the customer was coming back. For a lifestyle product, nobody went outside for a year.

Ben: Was the fact that you exclusively made performance running gear gave you the confidence to flip it back on? Because if you’re making all kinds of stuff in your factory and you’re pushing all kinds of stuff through retail channels, most of it is not going to sell, so you can’t actually open.

Jim: That’s right. Apparel and footwear inventory is life and death. You’ve got to manage inventory well. Because if you have too much, you’ll ruin the next cycle of inline product. Inventory is really critical, but we managed and played that cycle really well. We grew to 27% in 2020. We grew 31% in 2021. We would have been up 40% if not for supply chain.

Ben: What did you end up doing in revenue last year?

Jim: $1.13 billion. Great year. We cracked a billion. The billion dollar club is actually a rarefied club. There are probably maybe two dozen, global. Chinese brands are there now. It’s a great club to be in.

What makes us unique is it’s all premium, full price, full margin product. Most of the other brands have good, better, and best. Those are retail-driven merchandising strategies. They’re not really consumer-driven strategies.

Ben: Normally, we talk about seven powers as we drift into analysis here. You’re a Berkshire business, so we’re going to talk about moats. What is Brooks’ moat and how do you think about defending the castle now that you have what you’ve built?

Jim: We think a lot about it. I think there’s also something I’d add to that. Part of the moat can be business models. Business models can be really powerful. One of the things you can do as a company to create defensive moat structures is business model execution at scale.

We now are executing retail partnerships with the best retailers for running gear to runners at Super Jock ‘N Jill in Seattle, Fleet Feet running down in (I think) Menlo Park. Obviously, some of the better sporting goods players and outdoor from REI to Dick’s Sporting Goods, we’re their number one brand.

We’ve earned that over 20 years and we have deep, broad partnership programs with them. Digital marketing, consumer journey, runners are digitally savvy. They’re obviously all over the web. They start their shopping experience there.

We reach them in active evaluation mode. Once you start looking at shoes, if you don’t see our ad, I don’t know how we missed you. We’re spending a lot of money at runners now, maybe more money at people who run in active evaluation for running shoes than any other brand. Very focused. That’s not easy to do in our industry at scale.

I would say this is our moat. I think runnability, fit, feel, and ride, there’s a lot of good shoes out there. It’s actually not easy to make a great shoe. Anthony Fauci made a joke about shoes. “Vaccines are tough, they’re complicated. It’s not like making shoes.” We get a lot of that.

The refinement that goes into mile and making mile 26 acceptable, is really big. I think great product is not as common as you might think. The people on the inside, the frequent learners know. I think you always got to lead with product. That’s the first brand experience, product experience.

I think we do some hard things. We build a great product consistently, year in year out. It fits and it rides well. Then what we do on the retail side, partnering, activating in real life, running and selling shoes in real life events, and all the like, we do that better than anybody else. We service them. We deliver on time, complete. The digital piece, we’re excited about it. We’re still just getting started there, but we’re really focused on it.

David: I’m curious. I hadn’t even thought about Strava and the amount of data that you’re able to see from that. What does the digital side of running in the future look like for Brooks and for the industry?

Jim: It’s interesting because quantified self and those tools have been ubiquitous. They’re out there. The Apple Watch is a damn great product. What’s interesting about that is both Under Armour and ASICS have spent hundreds of millions of dollars on digital apps. I think they’ve really struggled a long time.

David: Runkeeper and MapMyRun both.

Jim. Exactly. I wanted to buy every one of those and Warren wanted me to do the multiple on EBITA. There was no EBITA. Let’s just say it’s hard to do acquisitions sometimes.

David: At least one of them was a completely free product, I think, right?

Jim: Oh, man. They don’t make money. Under Armour is trying to sort through that now. They’re starting to shrink, so as Adidas. Those tools are really powerful for data, but how do you monetize it? We haven’t gotten there yet, but we’re building a Brooks Run Club. Finally, we’ve launched.

It’s not a loyalty program, but we want to engage our zealots. We want to engage our true believers. The data piece of that is going to be key. We want to come up the kinetic chain and find a sensor system and a data capture system that can get to your biomechanics as you’re running. Because what happens is, if you run a marathon, your gait in the last 5–10 miles really degrades. And that’s where injuries happen.

We’re doing a lot. We have a lot of partnerships. We’re really trying to figure out how we get good runner data in real life, not just in the lab. In the lab, we can test everything, but we want to get out in the wild.

David: Do you think you need to do what the other folks in Oregon have done and build the whole consumer experience yourself? Is it a partnership?

Jim: We’re going to build it and we’re going to partner, too. Nike Plus is a fantastic ecosystem. It just is. I’d love to have an ecosystem like that. But we’re still selling more runners than they are.

We became the number one running shoe brand in the United States in the last 12 months last month, 21.5% share from performance running. We know where the battles are. I think one of those powers is we make money on that. The digital space, there’s a lot of carcasses there, but we’d love to have it, and we’re going to work on it…

…Ben: Yup. All right, one closing topic. You battled, survived, and beat cancer while building this incredible business. How has that changed your perspective on leading on the way you spend your days and on life broadly?

Jim: Let’s close it on a light note. Let’s talk about cancer. That’s the takeaway for these wonderful people. I didn’t expect it. It came out of nowhere. Unlucky. How did this happen? Esophageal cancer, I just felt awful. My worst running experiences I’ve ever had and I got the diagnosis. Chemo, radiation, surgery, complications in surgery, another surgery. but the good news is I’m cancer free. I think it’s gone. I think it’s out of my body. The bad news is I’m even slower and I’m kind of a Frankenstein in my systems, but it works. Everything works.

I think what I learned from that, though, is that every time I have a friend or a family member who gets cancer, I go to the web. You look at it, understand it, and what the treatments are. They always give you a five-year survival rate. My five-year survival rate was 20%, one in five. My five years is this November. Someone has kick its butt.

What I quickly figured out and I talked it through with my family and obviously with Warren, frankly, is that I decided that I was doing exactly what I wanted to be doing. I love what I’m doing. I’ve got family, I’ve got an active lifestyle, I’ve got this fabulous brand and a company that I’m a part of, and a team. I just love it. I don’t know what else I do, which is a problem.

I decided I didn’t want to live in fear. I didn’t want to live every day thinking about what I had to lose. I had a lot to lose. I didn’t want to be bitter about why me. I just decided I want to soak in everything I can on any given day. I want to be a CEO, I want to be a dad, I want to be a husband, I want to be a papa. I’ve got four grandkids. That was it.

I think for me, that was really powerful because I don’t want to be that cancer guy and they brought it up. It’s just not my thing. I’m glad to talk about it. I don’t hide it. I’ve learned a lot. I want to enjoy the things in life I really enjoy.

That’s where I learned, but I think everybody’s different. You do find out companies, when you hit challenges, you learn what you’re really all about. I think it’s the same for people, of course. I feel really lucky because I’m doing what I want to do. Cancer is in the rearview mirror. It’s good.

6. Martin Casado – The Past, Present, and Future of Digital Infrastructure – Patrick O’Shaughnessy and Martin Casado

[00:03:58] Patrick: How would you put chapter headers on the stages of cloud adoption, going back to, I think, Azure and AWS, are sort of mid-2000, 2005, 2006, thereabouts, relatively speaking, a short story. What do you think the major eras of the cloud story have been so far?

[00:04:15] Martin: Right before the cloud, of course, everybody ran their own internal IT. Right? And so they kind of write their own servers and their own wiring closets. The cloud showed up and the early usage was what you would typically find in a technology early adopter ecosystem. It’s more new projects and startups and hobbyists, the average workloads were relatively small. There was exceptions to that of course, like Netflix is a very famous one, which went all in the cloud very early. But in general, that was what it was. This is like 2005-2010 timeframe and still was very experimental. A lot of the time there was big discussions on whether the enterprise would actually go into the cloud. When I ran network and security for VMware, which is 2012-2016 timeframe, I think that was the more mainstream adoption of the cloud. You saw large organizations, traditional enterprise moving workloads to the cloud, very serious discussion with the Fed and the government. It became a mainstream way of doing things. If you were a large organization and you didn’t have a cloud strategy, I mean, you were either considered a laggard or a special case. That brings us to 2018-2019, and now we’re seeing a shift where the move to the cloud has implications on your finances, because now instead of you being able to buy a physical asset and internalize that, you’re basically paying a portion of your income to a third party.

Now there’s a lot of discussions around, how do you optimize the use of cloud? Is the right thing to go all in on cloud? Is it something that you do a portion or whatever? I just want to make one quick analogy, which is, I always view companies going in three stages, the product stage, the sales or growth stage, and then the operation stage. The product stage you’re finding product market fit. The sales stage is you’re getting to repeatable sales and growth. You don’t really worry too much about unit economics. And the operation stage is when you care about unit economics and you go into multiple products and you do all the operation of complex things. The cloud had gone through the exact same three phases, which first was trying to find product market fit, which tended to be within new projects, funding the projects. Then it went to the growth phase where everybody went all in and didn’t worry about the implications to the economics of business. Now we’re at the operations phase where we’re starting rationalize all of that.

[00:06:26] Patrick: Maybe tell the story of Dropbox, which I think as an individual company, is a great example of cloud isn’t just some panacea. It has incredible benefits in terms of how quickly you can get going, outsource the reliability to somebody else that’s just focused on this, AWS or whatever. But from a cost standpoint, it can get really out of hand. I think Dropbox is a good and probably unfamiliar to most tale of going the other direction.

[00:06:49] Martin: There’s basically two trends that happen at the same time. It’s important to understand those two trends to understand what happened at Dropbox and actually a number of other companies too, it’s not just Dropbox. The two trends are the following, the first one is cloud, which we talked about. The second trend is SaaS. And specifically what’s unique to SaaS is, is before if you were a software vendor, you would build software and you’d ship software, and somebody else would run it on their own infrastructure. Your COGS, your cost of goods as a software vendor did not include the infrastructure that it was being run on, because it was being run on somebody else’s infrastructure. For example, my startup, we built software for networking, we shipped it, other people would run it on their infrastructure. However, if your SaaS, if your product is software as a service, then part of your cost of goods is actually the infrastructure. Someone comes, says, “I’ve got a SaaS site and someone comes and uses it, then they pay me some, and then I pay say, AWS a portion of that.” That is a change of cost structure. The books look very different.

While the cloud is getting adopted, all software is going from basically on-prem to SaaS, and in some cases, and there’s many of these cases, it turned out that it was very tough to get software margins just because the cost of the cloud services on the backend was so high. The era of shipping software, we’d all say these companies have 80% margins because you basically write the code once and then it’s free to copy bits, so you just ship it to everybody else. Especially in infrastructure, there’s many companies that felt like they’re basically reselling a thin layer on top of AWS or one of the big clouds, and then paying a large portion back to them. For example, I know multiple companies that are household names, where they’ve got product lines that have 0% margins because all of the money goes back to the cloud services it’s hosted on. Dropbox very famously had this situation where S3, which is the storage layer on Amazon is not optimized for this use case of many small objects. They found that they were paying a tremendous amount. Now, they were a very large user of this specific use case. AWS was not optimized for it. They decided to build their own internal infrastructure and probably saved the company at the time, by moving off the cloud and taking it internally…

...[00:10:08] Patrick: There was a really interesting thing that you wrote about the interesting concept of lost market cap of companies that were big users of the public clouds. I’d love you to walk through that concept, because you mentioned maybe this saved Dropbox, the company, and I get that that’s a very special, specific case, but it sounds like there’s a bigger story here of lost margin and therefore lost market cap because of the use of public cloud. I’d love you to walk us through that.

[00:10:32] Martin: We did this analysis, a very simple analysis, which we said, “Okay, right now there’s a tremendous amount of money that SaaS companies spend on cloud.” Let’s say if they brought it inside and they were able to drop those costs by half, which most people agree that you can drop the costs by half by bringing it inside. If you could do that, what would that do to the stock price? Normally when people look at this problem, they say, “Well, if you bring this inside, yes, it’ll save you money. You’ll save 50%, but that money won’t cover the team, the complexity, because that’s not a lot of money.” But if you look at the leverage that increase in margin does to the stock price, now you can free up for a large company, potentially a lot of money, which will flow over to cash, so it could be a big win.

What we learnt is that we looked at just public software companies. We looked at 50 of them. We looked at all of their spend and we said, “Let’s assume you cut that spend in half.” Then we calculated their margins. And then we said, “Benchmarking against other public companies, if their margins were half, what would that do to the stock price?” It turned out that it would increase in aggregate the stock price by $200 billion. Just a tremendously high number. I think we wrote $100 billion to be conservative in the actual blog post, but $200 billion. That means if you’re a company that’s say, worth $10 billion, and you can reduce your COGS by a bit, you could now become worth $14 billion, and then you have access to that for debt and hiring or whatever else. Because those two trends happened at the same time you had the cloud trend, as well as the SaaS trend, I don’t think there had been a lot of focus on what it does to the margin structure. We did the first analysis and said, “Actually it’s huge and it can impact your stock price.” I do think, especially now in this market correction, it’s a good thing for companies to start looking at…

…[00:14:03] Patrick: Before we get to something like Kubernetes, a little bit more complicated of a topic, I’d love to just return to super basics around digital infrastructure in the first place. And maybe even go all the way back to the original AWS website, where I think it was storage, compute, database. You mentioned networking. What are the base level, most primitives of the digital world? What are the most important, big things that actually happen? Because I’d love to understand what’s changed in those areas, like compute sounds like compute. What is changing in those three, four, five base level areas?

[00:14:33] Martin: The traditional infrastructure’s computing and storage, and then databases. Prior to cloud, you’d buy a server from whatever, Dell or IBM or HP. You’d buy a switch from Juniper or Cisco. You’d buy a storage array from whoever, EMC. And databases from Oracle. So all those have now been, basically, collapsed into a software layer over basically merchant hardware in the cloud. So you can get the equivalent of just compute by TC2. You can get very flexible networking layers, where you can put security policies and that’s largely implemented in software within the cloud. And then you get these scalable services, like the database services that are scalable because they’re in the cloud. And so that’s the bread and butter of the cloud.

For a cloud is basically you take these traditional abstractions, compute and storage, that were connected to a box and now they’re just basically software services that you can spin up and they should be able to grow to the size of the workload. But what has also happened in the last say, five years is a number of services then built on top of those that are higher level abstractions. So for example, machine learning workflows, analytic workflows, different types of databases that focus on different types of query patterns. I want to do analytics, or I want to do LTP, or I want to do very fast queries or time series. We have seen this renaissance of infrastructure, again, which used to be tied to a box now being implemented as a software services in a way that’s much faster than we’ve seen historically for that exact reason. That it’s not confined to a box…

…[00:17:30] Patrick: How will that happen? It’s like up against a death star or something. Like facing these three big companies. What do you think the best entrepreneurs will do? Pick something like crazy specific and just go after a single thread? How do you think this innovation cycle will happen?

[00:17:43] Martin: All of these companies are like, very strong repeat founders and the companies are Mighty, Fly Out IO and Mosaic. So, what do these companies do? So Mighty is browser as a service. I don’t know about you, but right now even as we speak, I probably have 30 tabs in my browser. My laptop goes slow. If you use Mighty all of that’s offloaded and you get this crazy good experience, which is great for most of us, especially as the browser gets more workloads. What is Fly? Fly allows any developer to run compute workload at the CDN tier all across the world, which is important if you care about responsiveness to the users. And what is Mosaic? Mosaic is, basically machine learning as a service. So they provide the ability to run models very quickly for AI specific loads. So, what’s unique about all three of these companies is all of them are doing their own hardware. They’re looking to run servers, they’re racking and stacking. And these are very, very strong founders.

All of them are repeat founders and all of these companies have great traction. So what is happening here? I think it’s exactly what we’ve spoken about, which is there just are across the industry certain workloads that, if you look at that very specific workload, the cloud is just not optimized for them. And that provides room for the Mighty and Mosaics and Flies of the world to provide something that is a very attractive proof point or performance point or whatever it is, with respects to the clouds. And so I don’t think the answer is we’re going to see a lot of drop boxes, where the end customer builds their own data center. I do think we are seeing very concrete signs of third party companies coming in and providing cloud services that are just at a much better price point, or a much better performance point, or much more optimized for a workload. And because the cloud is growing to size, there’s enough market now for solvent companies to do these. And so I think this is the very beginning, again, of a much bigger trend.

[00:19:33] Patrick: Can you say a bit about your view of what I’ll call API first companies? Which I think a lot of people would include in this definition of digital infrastructure. If I can hire Stripe to be my payments processor by simply inserting a API into my software that I build and care about. And then there’s one of these APIs that’s proliferating for kind of everything. What do you see happening here? Is that infrastructure in your mind? Where does this fit into this equation?

[00:19:58] Martin: As markets grow, the unit to which you monetize gets more granular. And my favorite example of this, and it’s one that may be a cliche but it’s worth saying, is the car market. So, way back when in 1913, Ford had a factor called the Rouge River Factory. And this factory literally went in on one side, it was like water, rubber and coal. You know, and like iron ore, and what came out on the other side was cars. And the reason is there wasn’t a sufficiently large market for cars to actually have suppliers. You couldn’t be someone that provided wheels or whatever. And if you look at the car market now, I mean, there’s companies that provide nuts and bolts and you’ve got multiple tiers of OEMs and integrators, et cetera, et cetera.

So the same thing has happened to systems historically. So in the 1970s, the same company would build literally the chip, the motherboard, the sheet metal, the operating system of all the apps. And then of course the OS got disaggregated from the hardware and then the apps got disaggregated from the OS. So now what’s happening is the application itself is being disaggregated. You take any application, you blow it up and assume the market for this application or any application is so big that independent component of applications now can become companies.

So what does an application do? I mean, applications authenticate users, they need access controls, they need to send emails, they need to do payments. These are things that all applications can do. So it’s almost like every help or library in an application is now becoming a company. So much so that I remember even five years ago, you drive up 101, the heart of Silicon valley in the Bay area, and you’d have billboards where the entire company was an API. PubNub, Sendgrid, you know, Twilio. And so this is a major movement where now you don’t have to build a business app to build a company. And for an infrastructure person, this is super exciting because most of the founders I invest in are technical founders that are providing technical functions that are only useful to developers. And in the past, it was hard to build a business that way, but now you absolutely can.

If you’re in tech at all, or you’re an investor at all, I definitely think you should look at an application and assume that any sub-component does have the potential to now become a company, because the market is just so large.

[00:22:11] Patrick: What stage of that process do you think we are in? Twilio and Stripe, everyone knows turns out payments and sending messages. It’s almost like the equivalent of storage and compute in application building. Where do you think we are in that process?

[00:22:25] Martin: I think we’re still pretty early. I mean, on average, an application uses 17 external APIs. I think like a mobile app, something like that. But if you look at the use of libraries and open source and everything else, it’s still incredibly high for people having to integrate external components and management operate themselves. I think that there’s still a long way to go, especially as we get into kind of more complex things. So for example, every application often requires some sort of internal policy. Who can access what, or you know? And this is a very specific computer science problem. How do you build a language or a policy language that kind of accesses, that allows a third party to declare a set of rules and mitigates access to those rules? Like, this is a component in most programs that can be pulled out and turned into a company. There’s a number of companies looking at that, that are just getting started.

[00:23:16] Patrick: When it comes to this developer facing tooling, there’s this open source way of building and there’s the more proprietary, closed source way of building. What have you learned about what works well in which domain? And then I’d love to also learn, like if you’re an open source company versus not, what is more or less important as you think about product and go to market and everything like that?

[00:23:35] Martin: I’m starting to be of the opinion that as we move to SaaS and that’s the primary way of consuming infrastructure, which it seems to be, that open source matters a lot less. And the reason I say that is, if I’m a developer and I’m writing an application and I need to authenticate my users and I need to authorize their access to things, and I need to send them emails or send them SMS texts or whatever, I have two options. I could download some open source package and then operate that, or I could just use an API that somebody else operates. The secular trend is I’m going to use the API that somebody else operates. And if I’m doing that, whether or not the code for that is open source, doesn’t matter that much to me. So let’s take the case of it is open source. So, even if it is open source, there is some value there. A lot of actual code to running that service has to do with the operations of the service. Like, how do you make sure that it’s high availability? How do you debug it? How do you check for performance? Like, and that operations code is to be very specific to the actual service running. So it isn’t even useful.

So that would never be open source anyway. So even if I had the source code, I couldn’t really use it and operate it in the same way that somebody else could, or is running it. When it comes to dev tools, things that I am specifically using in my program as I develop, like that will always be open source and that’s very important. But anything that’s functional and offered as a service, I think the actual value of open source decreases. And what raises importance is actually open standards, which is, I still want be able to make sure that I’m not locked in to one and I can move between them, but that’s not an open source argument. That’s kind of an open standards argument. And so the role of open source has obviously shifted very, very quickly in the last 10 years, largely driven by this consumption with SaaS. And I think that we’re getting a more nuanced view of where it’s useful and where it’s not. Whereas 10 years ago, there was this broad consensus that open source is great and it’s going to take over the world. And that just doesn’t seem to be the case in the way that we all thought…

…[00:28:41] Patrick: Going back to this notion of, so if they’re the consumers of these APIs or little pieces of infrastructure, I absolutely love the Ford factory example, and what happens as it matures, that it’s so clean. What do you look for as an investor when you are seeing one of these, let’s say API forward or first companies for the first time? What is your method of investigation? How are you processing a new company?

[00:29:03] Martin: So throughout this all together, we talked about a trend. So there’s a lot of frontend developers. We talked about probably 100 to everyone backend. And those frontend developers, they’re building more and more of the application. So in the past, they had to… Were very tied to the backend more and more. Instead of having their own backend, they can use it an API from a third party company. Let’s say they’re using 20 little SMS or whatever. The interesting thing about these API companies that offer to the frontend is that the unit of consumption really is like a function call or an API call. So they almost have these consumer-like dynamics. So the primary evaluation criteria, to answer your question, and why it’s so different, in the past, if you’re going to evaluate a server company, who’s the buyer, what’s the go-to market motion, what’s the ACV.

You talked to a bunch of the buyers, you’d see if they can build the technology, et cetera. Now, with these API companies, you literally just can look at what the usage graphs are, how many users, how do they monetize them, et cetera, and it’s become much more of a bottoms-up, or SAS, or consumer type profile. So we stopped a lot of that approach to investing when evaluating these companies. It’s much less about can they build it, who’s the buyer, and it’s much more about how they use it in a practice, then it’d very interesting. A lot of these companies, they do. They’ve got these beautiful growth patterns, just like you’re looking at the next WhatsApp. They really are almost consumer-like phenomena.

[00:30:27] Patrick: What would be the most common red flags or disqualifying observations if you’re investigating one of these companies beyond lack of that nice looking usage or engagement?

[00:30:39] Martin: Well, I’ll tell you what I’ve gotten wrong. I do come from the older era where you actually evaluate the technology, you have a thesis on go to market. Often, we’ve seen these companies come in and they’ve got these beautiful usage graphs. They haven’t monetized yet, but we’re like, “Oh well, who’s going to pay for this?” Or this is just developers, like whatever. And then we kind of talk ourselves out of the deal, because we know the market better than the founder. And in almost every case, I’ve regretted that because the reality is, and this is an internal thesis of ours, is the graph in almost every case is just smarter than our theorying. The market actually knows what it wants.

These days, if one of these API companies is doing very well and the usage is great, I’ll give you an example, Hugging Face is a phenomenal company. And if you looked early on at the usage, this thing is a rocket ship, and you can have a bazillion theories why you can’t monetize the model, and you have a bazillion theories of why their go-to market is going to work. But the reality is the market loves it, it’s a great company. For me, it’s almost like a counter thing, which is, I do think that this API makes life a lot easier. You don’t have to have a grand unified theory about how things work, because you can literally just look at how this thing’s being consumed, because the consumptions become so bite-size; you get a lot of early signals. I think it really boils down to…

[00:31:53] Patrick: It comes down to usage.

[00:31:55] Martin: Yeah, to usage.

[00:31:56] Patrick: How should these things be priced? What have you learned about actually building the revenue model around something that looks more usage based? All these examples, AWS, what we started with, these API companies, they tend to be usage-based pricing. So what have you learned about that? Is that the right thing? Do you think that changes?

[00:32:12] Martin: It feels to me though, apps are for seat pricing, and infrastructure is usage pricing, and that’s basically how it is. And if you’re in the frontend, you’re not doing usage pricing, you better get there. You really have to. And if you’re apps, and you can get away from seat pricing, that just seems like that’s where you’ll end up. I do feel that when it comes to company building, there’s a few areas where there’s no simple answers. There’s a lot of stuff that’s systematic, like how do you hire your sales force, it’s pretty systematic. How do you create your org is systematic. But one of the things that’s just not systematic is pricing. Pricing is actually dictated by the shape of the market and the shape of the product. And it takes months to get it right. I’ll give you three mental landmarks, and I think the rest is just actual work.

So one of the mental landmarks is pricing is often fixed by the market. And so you should look at the ecosystem and the other types of companies and how they price and I think you should follow that model. For example, if you’re building on top of Snowflake, how Snowflake charges is going to be very similar to how the customer expects to buy. And if you’re building on top of that, you’re going to want to align with that. And I’ve been in many cases where the companies wanted to innovate on their own pricing model, but the ecosystem alignment just wasn’t there. And it was just painful until they had to change. I think another mental landmark is the market will tell you the price over time, but not initially. The less that you have public or the less that you force your opinion on, the better it is. I do think that a lot of early sales discussions is just to figure out pricing, that’s what it is. Your goal is to reverse engineer how they think about that. The good news is because the consumption is so much higher on these, and the unit of consumption is lower, it’s per API call, there’s kind of a lot of room to experiment…

…[00:40:14] Patrick: And as you think about the ways that all of this intersects with the real world now, which we really haven’t talked about. We basically talked about digital infrastructure that leads all the way up to applications at the top end, and the APIs in between and all this great stuff. But it seems also like as we mature, more of this technology will apply to the real world too, whether that’s new kinds of hardware, whether that’s intersection with physical goods like cars. How do you think about that side of things and maybe the hardware world of technology?

[00:40:45] Martin: People have a hard time grasping what, let’s say, AI and ML concretely provide, because it’s such a diluted buzzword. So for everybody that’s listening, the important thing to realize is that what modern AI and ML does, which we’ve never really been able to do before in systems, is take unstructured data, and digitize it, and add it to the typical logic of the program. And we’ve never been able to do this with vision, like objects out in the real world. We’ve never been able to do this with natural language in the level of ASCII we can. We’ve never been able to do this with voice or speech.

That technology just hasn’t existed and so we’ve never been able to build the big programs around them. And now we can, and it’s a sufficiently different workload that two things happen. First, it pushes software into the realm of the physical world. We can now see things and interact with things. And we’re talking quantum leaps of accuracy improvement. It also drives the type of hardware and software that we build, because the workload is so different, right? So we’re seeing tons of innovation all the way down to the ASIC level. Mosaic as a company is building a data center focused just on this type of stuff. So I think that this really is a massive impact on infrastructure at large, not just the infrastructure, but also what sorts of applications software can go after.

[00:42:10] Patrick: It’s very cool to consider what all that might mean. I mean like self-driving cars is like the obvious constant example of what computer vision might allow us to unlock. Obviously, cloud had this crazy impact on the services you consume. It’s unlocked innovation by reducing friction. As you think about what’s going on in the digital infrastructure world, period, what are you most excited about in terms of what it might unlock in the 2020s or over the next decade that maybe we’re just starting to think about?

[00:42:38] Martin: Any problem that human beings go after that’s been outside of the realm of software is currently in the realm of software. And this is farming, agriculture, oceanography, you name it. And so I am a tech optimist and tech maximalist. I think that part of our job is to solve problems. It has really been limited to IT, like information. And now I go from IT to just tech. You look at any industry, any industry at all, and I think that it’ll be touched by this. That’s, to me, just tremendously exciting. What’s interesting, I would just say very quickly is we’re still asking the question. Are these still software companies or something different? So you could say, this is just software going after agriculture. Now you still have a software business, or you could say this is still an agriculture business, or you could say it’s something totally different. That’s a question I’m personally very interested in…

…[00:45:34] Patrick: If you put your investor hat on, I guess your purely selfish investor hat, meaning you were just trying to maximize returns, and you could somehow have a crystal ball that would reveal some information about the future, which is currently uncertain, where if you knew what the future was going to hold, be super valuable to you as an investor, trying to earn a return. What would you ask of that crystal ball? Like what would you want to know about the future that you’re not sure which way it will go?

[00:45:58] Martin: I am very curious about where crypto lands, and I think there are three potential views, right? On one end, on the most negative and barren folks are like, “This is all fake. It’s just Ponzi scheme, yada yada, yada.” On the extreme other end, it’s a total reformation, not just of technology and companies, but an organization. This is like everything. You don’t just have routers. You’ve got crypto routers. You don’t have just storage. You have crypto storage. You don’t just have businesses. You’ve got dos. You don’t just have money. You’ve got DeFi, like everything changes. And then there’s a bit of a middle view somewhere. This is a continuum which says, “You know what? There’s something very innovative there on the ability to build networks. There’s a number of primitives that are very innovative on the ability to build applications. There’s a number of innovations on how you offer new services to consumers where you don’t know the endpoints. There’s a lot of great primitives, consumption to monetization layer, just like social was primarily a consumption monetization layer.”

In that future, that layer is on top of a lot of systems, but you still have traditional computer networking and storage. You still have traditional clouds. You still have to know all of those things. And it’s something that’s added to that. And I think the answer to that question given the amount of money that’s already involved is enormous. And I don’t think anybody knows the answer. I tend to be in the middle where I think that there’s a real innovation there. I think there’s real value. I think it’s a real unlock for a lot of new business applications and use cases. But I think that infrastructure itself, a lot of the traditional models still applies. You still have to build databases. You still have to use storage. You still have to understand the trade offs of asset. A lot of these things still apply.

[00:47:41] Patrick: Obviously distributed systems. Some of the smartest people in the world are working in distributed systems, not necessarily crypto networks, but just like the ability to distribute state or update state constantly faster, smoother, or whatever. As an infrastructure person, when you look at the current technology in crypto networks, maybe the dominant three or four, what are you watching or interested by or looking at, the consensus mechanisms, the scaling ability? What are the dimensions that you as an infrastructure person are keyed in on today?

[00:48:10] Martin: The crypto origin solves a very important problem. That’s traditionally not been solved practically. And that is allowing basically an anonymous set of people with no prior trust relationship to have strong guarantees on something, right? Originally it was a ledger, and then it’s become more to generalized compute. That’s a very, very real innovation thing. And that unlocks very, very interesting business use cases like we’ve mentioned. But distributed systems is one of those things that you just can’t paper over with a thin software layer. You can’t hide it under an API. You’ve never been able to. There’s entire languages that just help programmers manage distributed systems. What’s important is what developers end up using, or what distributed paradigms they end up using, because that will drive the capabilities of the system. So if everybody says, “This is a purely distributed world and everything I write must be purely distributed,” that will have some implications of the type of systems that you can build.

So the thing that I’m most interested in as kind of an old distributed systems guy is what are the nature of the applications? Is this going to land in the realm of purely distributed stuff? Is it only embarrassingly parallel applications, like DeFi is an embarrassingly parallel application? There’s other things that are embarrasingly parallel. Or, is this going to go more to the model of general compute? Is that something people are going to do? Are people going to build like the AWS in crypto? The answer to that is actually very, very significant, right? You could say, “Well, listen, traditional distributor systems are great for building AWS, and this is going to just be the consumption monetization layer, or it actually is going to cause innovation in the way that we do distributed programming in the future.” I don’t think that’s clear yet where that’s going to land.

7. Watch: How Does a Dead Fish Swim Upstream? – American Physical Society

Take a quick look at this trout swimming upstream. Notice anything unusual?

[Video of trout]

You’ve probably seen something similar countless times; the fish wriggles against the currents that push it backwards, slowly making headway until it turns and ducks out of the influence of the stream. Nothing special in that.

The only thing is, this particular fish is dead.

Yes, you read that right. No matter how lifelike it looks as it undulates across the tank, that same trout would just go belly-up if the current were switched off. So how can it possibly swim upstream?

A team of researchers from MIT and Harvard were equally surprised when they happened upon this phenomenon by accident. They’d been studying the way live trout conserve energy by swimming behind obstacles that block the current*, and unintentionally placed a dead fish in the experimental setup. When they took a closer look, they were stunned.

“It was incredible, very counterintuitive,” MIT researcher Michael Triantafyllou says, describing the shock he felt upon seeing the fish swimming upstream. He explains that while he knew trout were good at conserving and even extracting energy, he had no idea that they’d be able to extract enough energy from the surrounding fluid to swim upstream without expending any of their own energy. Immediately, the team started to investigate this new, seemingly impossible phenomenon.

As it turns out, objects that block the natural flow of water, like a rock or a boat, create a series of complex vortices in the current as the water navigates the obstacle. As anyone who’s tried to grab a fish knows, fish are quite flexible all down their spines, which allows the head and the tail to move independently of one another. In certain situations, the array of vortices forming behind an obstacle cause the body and tail to flap in resonance. This tilts the body in such a way that the vortices, which cause a pressure drop, apply a suction force that propels the fish forward.

As Triantafyllou explains, “You have a flow behind the obstacle, which creates a continuous stream of eddies. Each eddy contains energy and also causes the pressure in the fluid to drop… the eddy causes the body to flap back and forth, and the fish manages to extract energy.” Since all of the energy is supplied by the vortices, it doesn’t matter at all whether the fish is alive or dead, if the timing happens to be right.

---

Disclaimer: None of the information or analysis presented is intended to form the basis for any offer or recommendation. Of all the companies mentioned, we currently have a vested interest in Alphabet (parent of Google), Amazon, Meta Platforms (formerly known as Facebook), and Zoom Video Communications. Holdings are subject to change at any time.