Product-Market Fit - in the Realm of the Nation-State
Apologies for the delay, August has not been the relaxed holiday month I had envisioned. Seems deal flow is getting back to normal in this part of the world. More importantly, on to the content…
This week, we explore:
Product-Market fit in the realm of the nation-state
Why the AI-CCP tandem will replace the US-Open Internet tandem as the most influential technology-governance pairing of the 2020s
The accelerating pendulum of centralization and decentralization and why China’s AI-governance fit may be undercut later in the decade
AI, Crypto and the Narrow Corridor between anarchy and a Despotic Leviathan in Emerging Asia
I’ve been thinking recently about the compatibility between different governance structures and emerging technologies. Essentially, how easily technologies mesh with a nation’s governance structure or culture will foreshadow the breadth and depth of integration into the fabric of society. A broad implementation of a technology with good fit will then reinforce governance structures in a recurring flywheel. Technology-Governance Fit.
Peter Thiel managed to coin the crux of the insight in just six words: “Crypto is libertarian, AI is communist”.
While Peter’s distinction is ideological, ideologies are clearly tied to a nation’s political institutions and inform their actions. In a prior post on De-globalization & Customer Amortization, I touched on how US tech companies were not built in a vacuum, but benefited from the internet commercializing amidst a liberal US-led world order:
Charging forward, with a first-mover advantage, into the new global, digital frontier - alongside capital, supply-chains, people, information and institutions - all at the bequest and under the protection of a U.S. hegemon.
Open. Global. Free.
This was the era in which virtually all of the U.S. tech champions were birthed. In this era, blitzscaling flourished.
That world order is shifting.
The marriage between liberal, democratic values and a global, open internet have shaped the last three decades for billions of people. Individuals have grown accustomed to instant access to information, global social networks, and buying cheap goods from China online. Similarly, forecasting how the national pillars of ideology and governance react to or interact with new technologies like Artificial Intelligence, Crypto or CRISPR is essential to predicting the shape of our world beyond 2020.
The US and China present good illustrations not only because of their geopolitical weight, but because they are relative opposites. A young nation, liberal, individualistic and democratic vs. an ancient civilization, traditional, collectivist, and technocratic. Most countries tend to fall in one of the two spheres or along the messy gradient in between.
Artificial Intelligence & Fit
For example, AI seems compatible with top down, authoritarian regimes such as China. AI is enabled by massive data-sets collated in centralized databases used to train predictive algorithms. The rise of the smart-phone and the rapid digitization of China create a teeming sandbox for even a mediocre data scientist, let alone some of the world’s best. The government is paying attention. For the first time in history, centralized decisions may actually scale to allocate resources as efficiently as a market economy. AI is an essential tool to craft Xi Jingping’s 21st Century China. Beijing at the core, plugging into centralized platforms with massive distribution and fine-tuned algorithms to incentivize behaviors on a mass scale in line with the party’s vision.
Culturally, implementing broad (potentially intrusive) AI will face less resistance in China. With a century of war, revolution, and poverty fresh in the mind of older generations, privacy is a luxury which can be traded for stability and economic growth. The communist party has retained legitimacy by orchestrating perhaps the most staggering economic miracle in history since the 1980s. If you grew up in the chaos and poverty of the cultural revolution and a generation later you have stability, a healthy family, and can shop on a global bazaar with same day delivery paid for instantly via world class digital payments, privacy and free speech may seem like a fair trade. Chinese are simply more willing than their western counterparts to sacrifice privacy for convenience, individual freedoms for national stability.
Aside from this cultural practicality, the civilization is built on a long-standing tradition of respect for authority and hierarchy. Family structures and a collectivist mindset overshadow individual expression. Chinese are more willing to give up data when requested and, due to the tight-kit “guanxi” (roughly relationships), they are more comfortable sharing data in groups.
On the other hand, the American tradition has always had a keen suspicion of any centralized authority. Formed by migrants fleeing tyranny in their homeland, the founding fathers of the U.S. carefully designed a system of checks and balances in the early days of the Republic to ensure a diffusion of power. Liberal values champion freedom, individual rights, limits to power and democratic governance - a world where the people hold power over those that govern them. This liberal tradition seems incompatible with the political and economic impact of AI.
AI is a centralizing technology. In a world where data is the new oil, data network effects tend to form within a few large tech champions. More data = better algorithms = better products / services (think AMZN / BABA / SE recommendation engines) = more customers = more data. And around we go. The natural outcome is customers flocking to increasingly out-sized winners. And suppliers follow the customers. As highlighted by the recent US anti-trust hearings with Amazon, Apple, Facebook, and Google, the concentration of market power from this inherent data advantage strikes many as undemocratic and, therefore, unAmerican. To quote Chairman David Cicilline in the hearing:
Because concentrated economic power also leads to concentrated political power, this investigation also goes to the heart of whether we as a people govern ourselves, or whether we let ourselves be governed by private monopolies. American democracy has always been at war against monopoly power. Throughout our history, we’ve recognized that concentrated markets and concentrated political control are incompatible with democratic ideals.
Zooming out, Cicilline is saying: AI -> data network effects -> concentration of market power -> concentration of political power -> incompatible with democratic ideals. A squeamishly-flawed application of transitive property: “AI is incompatible with democratic ideals.”
Not completely, but certainly less compatible than a technocratic authoritarian regime in a largely collectivist and hierarchically society. While US politics have shifted from Jeffersonian to Hamiltonian over time, empowering the federal government, the distrust of centralized-authority - politically or economic - is alive and well.
I would change Peter’s quote slightly: “AI is a better fit for centralized governance; crypto for distributed”. As Hayek reminds us in the Road to Serfdom, Fascism and Communism are often discussed as opposites, but in reality, they are closer than they appear. Both aim to put the state in the central position of a society. Whether a radical left agenda, a radically nationalist one, or a monarch, AI is an excellent tool for the consolidation of power in the hands of few.
For better or worse, China’s combination of a top-down political organization, cultural acceptance of surveillance for “the greater good”, and a burgeoning hardware / IoT ecosystem proliferating from Shenzhen, mean that China will see a much broader roll-out of AI use-cases relative to democratic nations. Broader implementation and deeper integration mean more data, better algorithms, and more concentrated power. China and AI have technology-governance fit. In a bifurcating world which will likely be decided by economic leverage and cyber-security, this matters.
Crypto & Fit
On the other hand, opensource technologies like public blockchains tend to distribute power - appearing more aligned with democratic ideals.
Crypto empowers individuals to exchange without trust in a centralized intermediary - whether a bank, a big tech company or even a nation-state. Because of its decentralized nature and modern encryption techniques, these ledgers are very difficult to censor, hack, or edit. Not only can this technology take the form of money as we have seen with Bitcoin, but many developers are working on “Web 3.0” - a decentralized internet made from the bottom up with encrypted peer-to-peer networks in storage, networking and compute - the building blocks of the modern internet. Instead of relying on internet gatekeepers like AMZN, MSFT, GOOG, BABA, Tencent for our cloud infrastructure, why not tap into all of the spare storage / compute capacity on idling electronics globally, tied together by novel incentives and encryption? A distributed cloud.
And why stop at Big Tech? The very idea of crypto is subversive to the power of the nation-state itself; architected to live beyond its reach - in theory at least. While crypto’s libertarian bent makes those in Washington uncomfortable, it directly contradicts Beijing’s vision of the world. The great firewall was built to keep subversive ideas out. A fully decentralized and encrypted internet would be that much more difficult to contain.
In contrast to AI, crypto pushes power to the edges of the network.
As we forecast the intermingling of tech, ideology, and governance and its impact beyond 2020, its important to take stock of where we are today. History is often the best guide to the future and demonstrates an oscillation between centralized governance and decentralized governance in the modern era:
Monarchies / Feudal Society -> Aristocracy, Knights, Peasants (Centralized)
Post-Enlightenment Republics -> Jeffersonian, Small-land owners, Militias (Decentralized)
20th Century -> Mass production, Mass Warfare (Centralized)
1989+ -> Liberalism, Democracy, the World Wide Web (Decentralized)
~2015 - 2020+….
AI Giants (FAAMG, BAT), Surveillance Capitalism, Rise of China, Social Credit (Centralized)
Cryptocurrency, Web 3.0, 3D Printing, Opensource, Remote (Decentralized)
My own bet is the 2020s will be defined by AI and centralization. China’s arrival on the world stage, the increasing strength of Big Tech, rising populism / authoritarianism, the reversal of globalization, and cyber-balkanization all point to a world where nation-states / centralization of power is staging a comeback. A previously fragmented network of networks consolidating within a few champions. The advances in AI will provide this movement longevity - putting new tools in the hands of a few tech titans or government officials making them more difficult to challenge.
The competing distributed vision - blockchains - have struggled to scale and are still years behind AI in terms of application. In the meantime, BABA, Tencent, AMZN, MSFT, Google, Bytedance, Meituan and others are investing heavily in a centralized future. The technological architecture of gatekeepers and centralized databases is perfectly crafted for monopoly power economically or politically - depending on who might have access.
For these reasons, I see the AI-CCP tandem replacing the US-Open Internet package as the most influential technology-governance pairing of the next 10 years. This will influence the direction of emerging governments caught in middle.
However, it is unlikely to last. Like many cycles intertwined with technology, the pendulum swings are shortening. The feudal period lasted over half a millennium. Jeffersonian small farmers were replaced by factories much faster and the digital age has replaced the industrial age faster still. The decentralized era of the internet only lasted ~25 years before becoming dominated by a few gatekeepers. My sense is a serious challenge from a decentralized future will come faster - maybe the tail end of the 2020s. As Balaji Srinivasan pointed out in a 2013 Stanford lecture, industry development tends to follow the following S-curve:
With your daily phone calls, emails, texts, storage, eCommerce, enterprise tools, calendar invites, directions, and social directory - all concentrated in the same 5 companies - which now account for >20%+ of the S&P 500 - it would appear the existing web 2.0 is quite grown up.
However, the likelihood of the internet gatekeepers being dethroned is tied to the political jurisdiction in which they operate. While the US agencies are busy threatening to breakup big tech, China will promote its growth. While the US will condone decentralized innovation, the CCP has banned decentralized crypto-currencies outright - co-opting the movement with its centrally-issued DCEP. Using policy, the Chinese leadership may choose to stall the “Cambrian explosion of new entrants” instead preferring to solidify sprawling tech champions integrating ever deeper into civilian life. The CCP can then tap the tech company’s leadership or the back-end directly to further craft society in its vision. AI and centralization are a perfect match.
In my opinion, the more interesting question is how AI and blockchain are likely to interact once blockchains can properly scale. Will the centralization of data and casual-stance towards privacy no longer be an advantage when AI algorithms can leverage encrypted data-sets for training? Will a crypto-enabled world foster a robust exchange of secure, transparent data to break through data-siloes allowing for more accurate AI models than their centralized counterparts? Similar to how China leapfrogged the US in fintech because the US had invested in legacy infrastructure which was “good enough”, will the US be able to out-compete China in a world combining AI with decentralized exchange given its more open ideology?
I don’t know.
What I do know is its often the combination of technologies, the intersection of different fields and modes of thinking where true breakthroughs happen. While perfectly crafted for a world of centralized AI champions, will China be prepared for the wave that comes next?
Closer to Home - The Narrow Corridor
Given the breadth of governance structures in India + Southeast Asia spanning democracies, communist regimes, monarchies etc - I’ll leave it to the readers to parse through where their country sits on the spectrum. While this article focuses on strengths of technology-governance fit, often the tech on the opposite of the aisle is what is most necessary to strike a balance. A balance between anarchy and hierarchy. Strong leaders and a strong society. Centralization and decentralization.
Balance is key.
I have a liberal worldview. I grew up in a democracy and have a strong bias toward individual rights and freedoms. However, I also realize centralization and state capacity are crucial in the developing / industrializing phase of a nation. Relative to India, Indonesia, or the Philippines, Singapore is an interesting case study. A strong founding leader with a long-term vision for getting the nation off the ground with the stated goal of distributing power over time.
Perhaps the embrace of big data and AI can play a key role in the large, often hectic democracies of India and Southeast Asia to increase state capacity, inform legislation, and provide better public services. Strong leaders with the right tools will be essential to help craft a world where people can actually enjoy the freedoms allowed by their governing documents. The key is to make sure they do not over-step.
As history has shown, the line between anarchy and a Despotic Leviathan is awfully narrow. A delicate balance of power between the state and society.
AI is a tool to empower the state. Crypto a tool to empower society. Perhaps the real lesson is too much technology-governance fit in a single direction is not so optimal after all.