SITALWeek #371
Welcome to Stuff I Thought About Last Week, a personal collection of topics on tech, innovation, science, the digital economic transition, the finance industry, and whatever else made me think last week.
Click HERE to SIGN UP for SITALWeek’s Sunday Email.
In today’s post: AI is facing its first lawsuit in what could set precedence for many industries; Microsoft aims to replace office workers with AI, and they may succeed sooner than anyone thinks; a great article from Tom Junod on evolving human behavior as seen through the lens of one infamous NFL catch; can psychedelics also heal physical ailments?; prioritizing instinct before objective analysis; the many long-term influences on rates; looking back at a conversation on semiconductors two years later; and, why investing is like going to the movies.
Stuff about Innovation and Technology
Litigating AI
The first class-action lawsuit challenging the legality of a generative AI system has been filed against OpenAI, Microsoft, and GitHub over the Copilot software coding assistant. Copilot helps autofill snippets of code for programmers and was trained on large repositories of open-source code, which, despite being open for usage, is generally governed by licensing agreements that require users to provide proper citation. Given the eyebrow-raising copyright questions introduced by other types of generative AI services (like text-to-image engines such as Dall-E, also controlled by OpenAI), this case could set precedent for an explosively growing industry. The software engineer who is spearheading the lawsuit hopes that by bringing clarity to how these engines work and which prior art they rely on he can help accelerate their development rather than hinder them. Hopefully, we will have some useful rails sooner rather than later given the ecosystem’s already rapid consumerization and proliferation. As I’ve noted in the past (also, here), you may soon be able to just declare what you want a bit of software to do and generative AI will spit out an entire app. Toward that goal, Microsoft announced last week that Copilot is also adding a voice interface called Hey, GitHub!
Clippy Took My Job
Microsoft’s ultimate goal for OpenAI’s Copilot isn’t just to help coders, it’s to replace you. Well, perhaps they would argue they want to help you do your job more efficiently, but the writing is on the wall. This is a big Catch-22 for Microsoft because, as they make office workers more productive, they lose lucrative Office 365 customers. Effectively, if you let Copilot or other AI assistants into your workplace, they are going to learn what you do and largely replicate it. And, as jobs have gone remote, the task of tracking and replicating work has become even easier. This white-collar automation could have a dramatic impact on many middle-class jobs. Of course, this scenario isn’t anything new in the world of software. Kicking off the trend, ERP tools automated away many back-office jobs in the 1990s. However, the precedent and speed with which AI tools could begin automating the work of full-time office employees may catch the world by surprise (in addition to horizontal Office 365-based tools, AI assistants are also likely to develop for specific verticals). Amazon is perhaps one of the best examples of replacing human tasks with software. Their “Hands off the Wheel” program is described in this HBR article: “The transition to Hands off the Wheel wasn’t easy. The retail division employees were despondent at first, recognizing that their jobs were transforming. ‘It was a total change,’ [a] former employee…said. ‘Something that you were incentivized to do, now you’re being disincentivized to do.’ Yet in time, many saw the logic. ‘When we heard that ordering was going to be automated by algorithms, on the one hand, it’s like, “OK, what’s happening to my job?”’ another former employee, Elaine Kwon, told me. ‘On the other hand, you’re also not surprised. You’re like, “OK, as a business this makes sense.”’” The range of economic volatility will increase significantly as algorithms replace humans. For now, machines are only as smart as the data they are fed, and, effectively, they will all be fighting each other to maximize speed and profits. One example of the old “garbage in, garbage in” adage was Amazon’s SCOT system, which led the company to massively overbuild capacity during the pandemic. And, as I covered recently, automated yield-optimization algorithms drove a giant bubble in apartment rents. While the automation of white-collar jobs and decision making is inevitable, we should approach it with great caution and as much transparency as possible.
Miscellaneous Stuff
“We Stay for the Replay”
Tom Junod waxes philosophical for ESPN on the infamous Dez Bryant catch in the 2015 Cowboys-Packers playoff game, expounding on how humans have turned the concept of determining what’s real into the game of life itself:
“We invest endless faith in the power of technology to deliver clarity. But what it delivers is uncertainty, along with the prayer that better technology might yet yield better results...We watch football because the questions it requires us to answer are much easier than the questions required by politics and religion and law and science, not to mention real life. But the questions are increasingly becoming the same. How do we know what we know? How can we believe what we see? In football as in politics and in politics as in football, we come for the game; we stay for the replay. We watch the replay over and over, in the hope of resolution, but resolution is as hard to come by now as it was in the first instant replay, the one filmed by Abraham Zapruder in 1963. And that's why we have no choice but to keep on watching.”
Junod’s narrative is beautifully insightful and applicable to nearly everything, but it seems particularly germane to what has happened in the sphere of social media in the post-truth world. It’s an acutely accurate description, perhaps without intending to be, of the new Elon Musk era of Twitter.
Psychedelic Physical Healing
Nature photographer Jim Harris apparently cured his lingering, therapy-resistant paralysis, resulting from a horrific snowkiting accident that initially left him paralyzed from the chest down, with a single dose of psilocybin. Psychedelics are known to induce neuroplasticity and increase the effects of neurotransmitters – thus allowing neural regeneration – but our understanding of their effects and capabilities to promote physical healing is only rudimentary. It’s possible psilocybin reawakened dormant neural pathways in Harris’ paralyzed hamstring muscles, thus removing the last major impediment to his independent mobility. Outdoor Magazine quotes a researcher at a psychedelic company saying: “The changes in brain chemistry reverse atrophy and increase the neurons’ ability to rapidly repair damaged neurons, allowing them to begin their normal signaling process. Longer-term beneficial effects of psilocybin are believed to be related to regeneration of neurons and neuronal pathways that may have died.” Colorado recently became the second state to legalize medicinal psychedelics.
Stuff about Geopolitics, Economics, and the Finance Industry
Gut Before Brain?
A key step in our process of evaluating investments is the pre-mortem. As detailed in Time Travel to Make Better Decisions:
Pre-mortems help you determine what could go wrong before it happens. A pre-mortem is a way to try and picture yourself in the future and work backward to decisions made today. It’s similar in concept to Jeff Bezos’ regret minimization framework: “I wanted to project myself forward to age 80 and say, ‘Okay, now I'm looking back on my life. I want to have minimized the number of regrets I have.’” We do pre-mortems for every stock we consider investing in by transporting ourselves into the future and trying to guess at the answers in these scenarios: 1) We didn’t buy enough. Why? What questions/data would have clarified our understanding of the potential? And, 2) We should not have bought it. Why did we? What did we miss about the range of outcomes, the degree of predictions forced by the valuation, etc.? Similarly, if we are contemplating selling a stock, we try to answer these questions as our future selves: 1) We regretted selling it and ended up buying it back at a higher price. Why? And, 2) We never regretted selling. What negatives were there that we were right about? Often, the question isn’t about buying or selling outright, but getting to the truth of what position size an investment should be...This exercise may sound simplistic and obvious, but the key is to make time travel feel as real as possible to fully experience the thoughts and emotions of your future self. Making mistakes in investing (and life in general) is personal and painful – it’s a gut punch of regret. So, we try to literally vault ourselves into the future and see what it feels like to be selling a stock at a major loss – it’s a horrible feeling, how could we have avoided it? The answer can only be in the present. What information are we missing today, or, more likely, what questions are we failing to ask? What is it about the range of outcomes that we need to better grasp? Imagine you have an actual time machine to travel five years into the future. Imagine which path you took through time to get there and which ones you avoided.
Part of this process is trying to strike the right balance between rational analysis and intuition-informed decision making. The key is knowing how to properly weight objective data vs. subjective instincts, making sure to acknowledge that a large amount of our intelligence is collected and relayed by nerve cells distributed throughout our body, not just in our brain (see: Outsmarting Your Brain). Author and researcher Gary Klein suggests that there may be value in applying intuitive reasoning before rational deduction, lest your first-impression reaction be lost once you start digging into the details. That order is generally counter to the way most people inject a dose of intuition into decision making, with the gut feeling informing the final decision after all data have been analyzed. I’m eager to try this inversion with our pre-mortem analysis by considering the whole picture before precisely examining all of the hard data inputs. I think the ultimate process would probably be an iterative method that alternates intuition and objective analysis.
Patience and Interest Rates
Real interest rates have been declining for around 800 years. As we’ve previously noted, following the work of Ole Peters, this trend is most notably due to our enlarging debt burden existentially requiring lower rates for continued economic function. The current rapid rise in nominal rates is a Fed-policy based attempt to combat elevated inflation (itself a result of the last round of mistaken monetary and fiscal over stimulus), such that real rates (nominal interest rate minus the inflation rate) are actually still not that far off the 800-year downward slide. Of course, that is of little consequence if higher rates collapse asset values and wipe out debt. Since one person’s debt is another person’s asset, when we wipe out those assets, the consequences are far reaching and long term. There is an elevated chance that the current monetary policy intervention will inadvertently stop the centuries-long trend in rates for a long period of time by destroying asset values too quickly for the economy to recover. The economy is a complex adaptive system, defined by resilience and adaptability, but any complex ecosystem ravaged by overwhelming external factors risks breakdown and extinction. My base case remains that the economy is largely self-healing, and I’ve argued that the speed of information in modern times increases the ability of the economy to solve many of its own problems. But, that assumes governmental meddling doesn’t create new difficulties not easily corrected via normal market mechanisms. One interpretation of the rather extreme volatility in the markets lately is investors grappling with the unknown outcome of whether or not the Fed’s interference will prevent the economy from self-healing, and in doing so cause an amplified recession, or worse. Time will tell.
As Peters has pointed out, a consequence of ever-rising debt levels is rising wealth concentration. It’s possible that a little bit higher inflation and rates than we have seen for the last decade are actually a positive indication that redistribution and growth are happening in the context of somewhat slower debt expansion (see Blueprint for Rebooting Distributive Era). As a side note, in addition to the mathematical reasons that ever increasing debt is symbiotic with lower rates, economists at the University of St. Andrews recently updated a working paper hypothesizing that another reason for 800 years of declining rates is selective breeding for patience. As I am fond of saying, the first species that humans domesticated was not dogs but, rather, ourselves, thanks to our constant attempts to improve the quality of our tribes and interpersonal interactions. There is a bit of mental gymnastics to understand the theory that enhanced patience is driving lower rates, but the main point I would make is that there may be multiple paths to explaining the gravity pulling interest rates down over the long term as a civilization grows in size, abundance, and perhaps even in patience. Regardless of whether the patience theory has explanatory power, our hope for the future is heavily tied to our desire to borrow and invest in things like families and businesses. There may be good reasons to analyze how the prevailing economic backdrop impacts long-term rates and how the Information Age might be different than the Industrial Age. It will be interesting to see whether or not the TikTokification of life and declining birth rates (which could imply diminishing patience and declining hope) are enduring trends that will impact interest rates. Maybe someday we will come to learn that interest rates rise as our attention spans shrink.
Shifting Winds Plot New Economic Course
It's been two years since Jon and Brinton talked about the semiconductor ecosystem with Shane Parrish. Toward the second half of the interview, the conversation turned away from the handful of companies that have come to rule the fate of the world and toward geopolitics, including the potential for the US to impact China's semiconductor ambitions by banning capital equipment sales. It has been clear to us for several years that the geopolitical range of outcomes for the world continues to widen – and access to leading-edge semiconductors sits at the center of the conflict. For example, we thought the range of outcomes for Apple was widening due to a strong dependency on China for manufacturing. While Apple has sought to diversify their supply chain in the past several years, they remain fragile to China. And, sweeping restrictions from the Biden administration put in place in early October appear much harsher than policies implemented by the Trump administration. More broadly, since that interview, we've seen baby steps to diversify the semiconductor supply chain away from China and Taiwan. TSMC is building fabs in both Japan and Arizona. Intel has announced they're getting into the foundry game with a big investment in new Ohio-based fabs. It seems to us that we are in for a decade-plus of moving from geographically centralized to decentralized production, which could spark higher levels of capital investment than we've seen in the past. This is not just true for chips, but for every industry, as thirty years of globalization, which peaked a decade ago, gradually unwinds (to varying degrees). Focusing on whether or not the range of outcomes is narrowing or widening around a given scenario allows us to adjust our thinking and ultimately adapt to the changing environment.
Today, the range of outcomes for nearly every aspect of the global economy is widening as decades of globalization, accommodative monetary policy, and population growth transition to deglobalization, a scarcity of capital, and a rapidly aging population. This will result in capital allocated to new areas with the highest potential returns, resulting in a whole new set of long-term economic drivers with profound global impact. We will be writing more about these seismic changes in the coming year, and we touched on the potential range of outcomes in our recent quarterly letter, excerpted here:
The third quarter was marked by the ongoing tug of war between investors, policy makers, dictators, and the painful landing of the pandemic-inflated economy. In the short term we expect uncertainty to persist; however, over the long term we know that the future, like always, will be determined by the optimists. Investing, like nearly everything in life, is a form of storytelling. When anyone buys or sells a stock, they are crafting a story about the future. But, of course, no one can predict the future with any precision, and relying on the past can be just as problematic. Therefore, the stories we tell must be based on what we can clearly discern in the present, no matter how foggy the environs might seem. We then live through these stories as they play out in real life, fact-checking our narrative with reality to identify plot holes, inconsistencies, and elements of truth that help us refine our story. There are some types of plots that tend to play out more often than others, there are always unexpected twists and, if you are lucky, a deus ex machina. At NZS Capital, we think stories of optimism – where the protagonist is adaptable and creating more value than they take (non-zero sumness) – occur much more often than stories of cynicism and pessimism, where the characters are rigid in their beliefs and extract too much value from society. The market, however, tends to be more cynical than optimistic, which creates cycles of fear.
Investing in the stock market is like buying a ticket to a movie about which we have little a priori knowledge. We can make educated guesses based on the title and trailer, but sometimes a seemingly feel-good rom-com has an unexpected horror subplot. The current market volatility implies a tension between two completely different plots for how our economic future will unfold. One story is a doomsday movie wherein government policy makers, in an attempt to rewrite their “certified-rotten” pandemic-policy script, send the economy careening off a cliff while geopolitical tensions send humanity into nuclear armageddon. Herein, there is no hero, and the only way out is time and patience. Like Major King riding the bomb in Dr. Strangelove, this movie ends with more uncertainty and questions than with which it began. The other script, however, has a happier ending. The underpinning of this second narrative is the self-healing ability of the economy, with innovation and hope driving a long cycle of post-WWII-like prosperity, investment in green energy, and a steady increase in global economic resilience. The first story implies that the last forty years of globalization, low interest rates, deflationary forces, and rising inequality will be met with a long and harsh punishment. In contrast, the second story plays to the evolutionarily ingrained resilience and ingenuity of humans. Of course, there are thousands of ways the future could play out, and we won’t know the ending until the lights come up and the curtains close.
Like the complex world around us, the global economy is dominated by power laws and extremes. Humans, however, tend to be linear thinkers, so we can miss emergent, game-changing events unless we train ourselves to look outside our typical mental confines to routinely scan left field and beyond. Some of the best stories of the future may be written in surprising locations and combine different topics and technologies in novel ways. Importantly, in times of volatility, policy changes, and technological disruption, economic resources often shift and refocus on newly emerging areas. A series of events, such as the world has recently experienced, can create pivot points for society. We continue to see significant opportunities in the engines of the analog-to-digital transition of the global economy, such as software, semiconductors, and the Internet. But, we expect resources and attention to shift to new origins of asymmetry in areas like AI, automation, healthcare, and other parts of the economy. While we cannot know the precise storyline for what lies ahead, we can appropriately prepare for all potential plot twists.
✌️-Brad
Disclaimers:
The content of this newsletter is my personal opinion as of the date published and is subject to change without notice and may not reflect the opinion of NZS Capital, LLC. This newsletter is an informal gathering of topics I’ve recently read and thought about. I will sometimes state things in the newsletter that contradict my own views in order to provoke debate. Often I try to make jokes, and they aren’t very funny – sorry.
I may include links to third-party websites as a convenience, and the inclusion of such links does not imply any endorsement, approval, investigation, verification or monitoring by NZS Capital, LLC. If you choose to visit the linked sites, you do so at your own risk, and you will be subject to such sites' terms of use and privacy policies, over which NZS Capital, LLC has no control. In no event will NZS Capital, LLC be responsible for any information or content within the linked sites or your use of the linked sites.
Nothing in this newsletter should be construed as investment advice. The information contained herein is only as current as of the date indicated and may be superseded by subsequent market events or for other reasons. There is no guarantee that the information supplied is accurate, complete, or timely. Past performance is not a guarantee of future results.
Investing involves risk, including the possible loss of principal and fluctuation of value. Nothing contained in this newsletter is an offer to sell or solicit any investment services or securities. Initial Public Offerings (IPOs) are highly speculative investments and may be subject to lower liquidity and greater volatility. Special risks associated with IPOs include limited operating history, unseasoned trading, high turnover and non-repeatable performance.