Friends, 

A couple of months ago, my friend Ryan Pripstein introduced me to the concept of “productive bubbles.” I haven’t been able to stop thinking about it since. The core concept is simple: throughout economic history there have been examples of rampant speculation that created some of our most important infrastructure including railroads, widespread electricity, and the internet.

How can we look at our present world through this lens? Today’s piece explores the concept and its connection to recent venture capital, crypto, AI, and climate frenzies. My hope is that it will offer a new perspective through which to view tech’s never-ending, never-dull gyrations and innovations. 

Why should we thank Elon Musk? What crypto inventions might stand the test of time? Should we worry about over-investing in AI? We’ll explore all of these questions in today’s piece, below. 

Jump in

👋 Before we get into the piece, I also wanted to extend a very warm welcome to the new readers who’ve hopped aboard over the past few weeks. You’ve joined an incredible crew of more than 78,000 investors, builders, operators, and thinkers.

In the weeks and months ahead, you can expect a combination of in-depth case studies, noteworthy interviews, company watchlists, big questions, and explorations at the frontier of AI, venture capital, and beyond. Here are five recent pieces you might enjoy:

1. Case study: The Remaking of Kleiner Perkins

2. Interview: Reid Hoffman on generative AI, nation-states and friendship

3. Watchlist: 10 AI startups investors and founders are monitoring

4. Case study: Substack, an Empire of Narratives

5. Big question: Who becomes an entrepreneur?

It’s amazing to have you here, and I’m looking forward to learning alongside you.

Brought to you by Mercury

Secure banking, the startup way

Every founder’s top priority right now? Keeping their cash safe. But that doesn’t mean they have to settle for a bank with clunky features and complicated services. Mercury offers secure banking* engineered precisely for the pace and creativity of startups — all through an intuitive product experience that innovates alongside you. 

Via Mercury’s partner banks and their sweep networks, customers can access up to $5M in FDIC insurance — 20x the per bank limit. These sweep networks protect your deposits by spreading them across multiple banks, limiting the risk of any single point of failure — and giving you peace of mind. 

You want to spend time actually building your startup — not worrying if your cash is safe, waiting in line, or navigating a confusing website. Plus, with Mercury, it’s simple to get started. Applying takes minutes and many customers are approved and onboarded in less than two hours. 

Join more than 100,000 startups that trust Mercury for banking.

Learn more

*Mercury is a financial technology company, not a bank. Banking services provided by Choice Financial Group and Evolve Bank & Trust®; Members FDIC.

Exploring Productive Bubbles

Actionable insights

If you only have a few minutes to spare, here’s what investors, operators, and founders should know about “productive bubbles.”

• Beneficial bubbles. Some speculation may be worth the pain. According to economist and investor Bill Janeway, “productive bubbles” have helped create some of our most important infrastructure, from railroads to the internet. While plenty of waste occurs during these frenzies, they improve life for successive generations. 

• Uncertain applications. Part of the reason even productive bubbles are wasteful is that it’s incredibly difficult to find “killer applications” for novel technology. For example, one early use case for the telephone was listening to music; meanwhile, radios only embraced broadcast after World War I. It may take decades for innovation to find its ideal implementation. 

• The limits of foresight. Janeway’s theory demonstrates that even when an investor is correct about the world-changing potential of a new technology, they can still lose their shirt. Plenty of investors that shrewdly foresaw the impact of the internet, for example, were still decimated by the dotcom bubble’s bursting.

• Useful remnants. We are emerging from an “everything bubble,” that touched public markets, venture capital, crypto, and beyond. A new AI froth is forming rapidly. What, if anything, are the productive remnants of our past and present immoderacies?  

• Climate next? Though Elon Musk and Tesla helped create a green tech bubble, galvanizing the creation of new electric vehicles, a broader climate bubble could be on the horizon. Given the existential stakes, we might all benefit from a period of frenzied over-investment in the sector.

Sometimes, the economy really does hang by a thread. 

On May 3, 1893, the National Cordage Company – a maker of twine and rope – collapsed, sparking the meltdown of the American economy and puncturing the era’s biggest bubble: railroads. 

In our age of skimming, soundless electric vehicles, darting scooters, and slaloming segways, the railroad sounds like a dusty inheritance: a grime-dappled silver platter, a sunworn sofa set. Not so in the 19th century. “The one moral, the one remedy for every evil, social, political, financial, and industrial, the one immediate vital need of the entire Republic, is the Pacific Railroad,” the Rocky Mountain News raved in 1866. This was the technology of the future, a galloping machine that ate fire, breathed vapor, and folded the American map until its cities touched.

What should a financier do when confronted by such ingenuity, such startling modernity? Invest, of course. Don’t concern yourself with petty things like prices or projections; overlook rhapsodic marketing and overlapping lines; put your money into the designs of the Philadelphia and Reading Rail Road (P&R), Northern Pacific, Union Pacific, and dozens of other scrabbling tracklayers. These businesses were the startups and unicorns of their day: fast-growth “sure bets,” reaching toward the horizon. Indeed, in 1871, P&R became by some measures the most valuable company in the world, valued at $170 million, north of $3 billion today. 

Flush with cash, powered by low-cost immigrant labor, boosted by government support, and unmolested by oversight, “railroad fever” defined much of the 19th century. An “orgy” of construction saw miles of track increase from 45,000 in 1871 to 220,000 by 1900, a nearly five-fold increase. In the 1880s, railroad firms made up an astonishing 80% of the New York Stock Exchange’s listings and delivered twice the government’s revenue.

And then, it all came crashing down. Panic over gold reserves and the demise of the P&R fretted investors, but the sudden collapse of National Cordage marked the true beginning of the “Panic of 1893.” In the following months, thousands of railroad workers lost their jobs as Northern Pacific and Union Pacific followed P&R into bankruptcy. By the next year, 25% of railroad businesses had made the same bleak journey. A broader economic depression lasted years. 

Time reveals truths obscured by the present. While many suffered during that spell – losing work and watching as wealth evaporated – the railroad mania of the 1800s merited the damage it caused. Though wild and unhinged, speculation attracted massive pools of capital for a monumental infrastructure project. Companies failed, and investors lost their money, but, for the first time, tracks knitted American industry together. This sparked new efficiencies and spawned radical industries, including mail-order retailers like Sears.

It was a perfect example of a “productive bubble,” a term coined by economist and investor Bill Janeway. Unlike pointless fits of speculation such as the Dutch tulip craze of the 1630s, “productive bubbles” leave society better off, undergirded by new technological infrastructure, even as investors lose their shirts.

No imagination is needed to understand how Janeway’s concept fits into our current era. We have been living between bubbles of different sizes and maturations for much of the last decade and a half, riding the froth and foam of venture capital, crypto, meme stocks, SPACs, and, presently, AI. Some have burst or are in the process of doing so, while others are still inflating, amassing hot air.

Which, if any, of our speculations will prove most productive? Which frenzy might future generations appreciate? 

Productive bubbles

Before looking at the ruptures of our contemporary era, it’s worth fully defining Janeway’s concept. 

Economic history is littered with bubbles. Indeed, a bit of froth is liable to form wherever a market emerges. Over the centuries, gimlet-eyed speculators have salivated over gold, silver, uranium, real estate, equities, digital tokens, flowers, and beyond. 

Typically, these events are viewed with a mixture of disdain and disgust. How could so many be so foolish for so long? What derangement excuses such total, brainless waste? In this rendering, virtually everyone loses; there are no heroes. The world is awash with charlatans, opportunists, and those feeble-minded enough to believe both.

Sometimes, this is the whole story. Sometimes, a bubble offers nothing more than the nausea of a sharp rise and stomach-churning drop. The tulip craze of the 1630s, the Beanie Baby mayhem of the 1990s, and most of the ICO psychosis of 2017 fit this bill. These speculations hoovered up capital and delivered next to nothing of value, cleaning out investors in the process. 

As Bill Janeway chronicles in his prodigiously named book, Doing Capitalism in the Innovation Economy: Reconfiguring the Three-Player Game between Markets, Speculators and the State, not all bubbles fit this profile. The annals of financial exuberance include many events in which hysteria produces something profoundly valuable. Speculation has sponsored critical infrastructure, created only thanks to “productive bubbles” that supply the capital needed to bring them to fruition. A selection of great American bubbles: 

• Telegraph fever (1840s-1850s). Samuel Morse’s new communication device sparked widespread excitement and investment. The number of telegraph miles rose ten-fold between 1846 and 1852 – never mind that many were redundant or effectively unusable. Though most telegraph firms failed, America benefited from new informational infrastructure and applications – like sending money by wire – built atop it.

• The electrification boom (1920s). Favorable economic conditions, governmental intervention, and financial speculation expanded the prevalence of electricity in America. At the start of the decade, 35% of households had electricity; within nine years, it had nearly doubled to 68%. The crash of 1929 pummeled the economy, including utility providers, but America’s electrical availability had been decisively enhanced. 

• Internet bubble (1990s). During the 1990s, more than $30 billion was spent building 90 million miles of fiber-optic cable. Meanwhile, a generation of “dotcom” businesses snaffled capital and reached wild valuations. In 2000, the bubble burst, resulting in the culling of tech companies and infrastructure providers. By some estimates, just 5% of fiber-optic capacity was used in 2001. Of course, the decades since have demonstrated the value and importance of the internet and the infrastructure that powers it. 

Ours is a history littered with productive economic confusion. Though these examples hopefully bring Janeway’s core concept into closer view, questions nevertheless remain. Among them:

1. How do you identify a productive bubble?

2. Can innovation occur without mass speculation?

3. How do companies forged in a bubble fare?

Firstly, how do you identify a “productive bubble?” Time tends to reveal whether our deliriums have been well-directed. However, a more concrete assessment method is to see if speculation centers around a “general purpose technology” or GPT (not that kind). As defined by the economist Timothy Bresnehan, a GPT is “widely used, capable of ongoing technical improvement, and enables innovation in application sectors.” Bresnehan adds that the GPT and application positively amplify one another; improvements to the GPT empower new applications to form, and new applications demonstrate how the GPT can be better developed. 

Janeway cites steam power, electricity, and the internet as exemplar GPTs, relied upon by “application sectors” like railroads, manufacturing, smartphones, and search. Bubbles that facilitate the exploration and expansion of GPTs and associated applications can be productive. 

Secondly, could this infrastructure have been built without all the speculation? It’s a reasonable question. Even when productive, bubbles are messy affairs that leave many businesses bankrupt. As Janeway writes, “It took the wastage of a bubble to fund the exploration that would yield Amazon and eBay and Google.” Despite its profligacy, however, financial speculation can gather the large sums of capital needed to build new networks. Janeway suggests that only the state can mobilize equivalent resources.

The state and speculator’s involvement in such activity is “decoupled from the rational calculation of gain from [a] project over its economic life.” The government does not build a highway system because it seeks a financial return; it does so for the sake of its citizenry. For example, in France, the government devised and implemented the railway system – the result is a much more orderly network without the overlapping redundancies that the American capitalist iteration involved.

Equally, the financier that YOLO’d into Union Pacific rationally could not have expected the project’s cash flows to repay his investment. Indeed, the more novel the innovation, the more unknowable its realistic cashflows. Rather, the financier was driven by a desire to “[ride] the psychology of the market,” making money as the asset bubble inflates.

Innovations do occur beyond state programs and financial frenzies, of course. Research and development (R&D) centers like Bell Labs and Xerox Parc have a rich history of invention. Janeway notes that none of the networks that emerged in the wake of the railroad excesses required as much capital, with less money needed to roll out local telephone networks. Though not the only way, bubbles do amass considerable capital around specific assets – often at scales not possible by other means.  

Finally, how do companies founded in bubbles fare? Though not a comprehensive historical study, one research paper suggests that bubbles may produce more highly valued businesses, albeit at a cost. “Investment cycles and startup innovation,” written by Ramana Nanda and Matthew Rhodes-Kropf of Harvard Business School and the National Bureau of Economic Research, finds that startups that first raise capital in a frothy environment are more likely to file for bankruptcy, while those that eventually go public do so at greater valuations, with more and better-cited patents. Given the power law dynamic of venture capital, that skew can prove attractive. After all, history will not remember John Doerr’s dotcom misses, only that he backed both Google and Amazon.

Keep reading

Puzzler

If you’re stumped, just respond to this email for a hint.

You’ve heard me before, yet you hear me again. Then I die 'til you call me again. What am I?

Last week’s winners were Bruce G, Dan M, Krishna N, Joshua K, Eugene H, James A, Thomas B, and Lim Cheng K. All found the correct answer to this puzzler: 

What's black when you get it, red when you use it, and white when you're done with it?

The solution? Charcoal. Well played, all. 

Until next time,

Mario