Centralization is the Root of All Evil
“Fix the money, fix the world.”
One day you will pay an AI a small fraction of a satoshi to transmit this article to your device…
“Fix the money, fix the world.”
Bitcoiners are fond of repeating the adage “Fix the money, fix the world.”
We believe in this pithy adage but we also recognize that it compresses decades of change, growth, and disruption into the space between its clauses. We want to explore this space, to understand what fixes will occur, and in what sequence, as bitcoin matures over the decade ahead and far beyond.
What is wrong with money, that it requires fixing? The usual answers are that fiat currencies are prone to inflation, surveillance, censorship, and seizure. We believe these problems, though important, are merely symptoms of the root problem: centralization. Inflation happens because governments and central banks control monetary policy. Surveillance and seizure are possible because a relatively small number of banks and fintech companies custody and control access to money. To quote Satoshi, the root problem with conventional currency is all the trust that’s required to make it work.
What is wrong with the world, that it requires fixing? We believe it’s the same problem. The requirement to trust centralized public and private actors is not limited to money. All sectors of our economy, from primary and secondary industries that extract resources and turn them into goods, to the higher layers which provide services and knowledge to us all, have become increasingly dominated by a small number of global corporations. Governments have grown larger and more powerful through purchasing the goods and services these corporations provide and, as a result, have grown more reliant on and captured by them.
Despite prominent mid-century warnings, we failed to dismantle the military-industrial complex. Instead we grew new industrial complexes intertwining government with finance, pharmaceuticals, prisons, healthcare, non-profits, technology, universities and more. We don’t need to fear AI but perhaps we should fear the coming AI-industrial complex.
As centralization worsens, competition decreases, regulatory capture becomes easier, enshittification accelerates, corruption and fraud proliferate, and the interests of central actors and the people they serve diverge ever further. Yet centralization continues to metastasize because the benefits of centralization have historically outweighed our collective concerns over its dangers and failings. We didn’t explicitly choose centralization, it simply crept up on us by being incrementally better and stronger and more resilient, time and time again.
It is easier and cheaper to live in our comfortable panopticon than to live off-grid in privacy and hardship, to purchase mass-market planned obsolescence rather than bespoke durability, to get killer deals online and thereby kill your downtown shopping district, to let Google read your email rather than run your own email server. But this has consequences!
Now centralization is entrenched, not only in our institutions and infrastructure but in our discourse, in our very minds. Our politics is a sponsored debate over which centralized actors we should trust, rather than any attempt to limit or reverse centralization itself.
There seems to be no alternative model around which to rally. How else, besides governments and armies, besides banks and stock exchanges, besides public companies and NGOs, besides regulators and agencies and courts, besides intellectual property protections and tax exemptions can we incentivize, protect, and deliver this complex thing we call society at our global scale?
2 companies control 42% of the global seed market
4 companies control 61% of global pesticide sales
OPEC produces ~35% of global oil
1 company (TSMC) controls ~67% of chip-foundry revenue
5 chipmakers control ~90% of advanced manufacturing capacity
China produces ~75–85% of global lithium-ion batteries
Amazon drives ~36% of U.S. e-commerce
4 airlines control ~80% of U.S. air travel
3 firms manage ~80% of U.S. prescriptions
3 companies control ~72% of digital advertising
3 firms control ~63% of cloud infrastructure
5 publishers produce ~57% of academic research
Nvidia controls ~92% of AI data-center GPUs
ChatGPT holds 70%+ of the chatbot market
3 companies power ~77% of enterprise AI usage
Central governments collect ~53% of OECD public revenue
5 countries hold 5/15 of UN Security Council seats and veto power over global security decisions
5 contractors win ~63% of US defense contracts
A market, not a protocol.
Bitcoin is both the soundest money in history as well as the most successful decentralized network on the Internet. These accomplishments are not independent. They are inseparable.
Bitcoin is sound money because it is inflation-proof and resists seizure and censorship. These desirable monetary properties are delivered by a decentralized peer-to-peer network. This network is successful because it is highly available and resistant to spam and capture. These desirable network properties are created by paying people to operate the network. Bitcoin is money, but it is also a market with an order book, a ledger, and ongoing price discovery. Through this market, bitcoin users directly pay bitcoin miners to produce blocks and settle bitcoin transactions.
Peer-to-peer markets combine the pricing, allocation, and optimization powers of centralized markets with the resilience, censorship-resistance, and privacy of decentralized peer-to-peer networks.
A market where peers directly exchange services for money is a new idea in the history of online, peer-to-peer networks. Such a market is only possible because bitcoin is money, not file-sharing or packet-routing. Being money allowed bitcoin to evolve past the narrow niches occupied by traditional peer-to-peer networks into something more general and far better-adapted for the scale of the modern Internet: a peer-to-peer market.
Peer-to-peer markets combine the pricing, allocation, and optimization powers of centralized markets with the resilience, censorship-resistance, and privacy of decentralized peer-to-peer networks. Peer-to-peer markets are powerful tools for scaling decentralization online because the direct exchange of money for services creates economic (dis)incentives that limit adversarial behavior and promote cooperative behavior. Traditional peer-to-peer networks, not being markets, must instead rely on porous filtering mechanisms to limit adversarial behavior and good will to encourage cooperative behavior. This does not scale, and traditional peer-to-peer networks are plagued by spam, inefficiency, and brittleness.
Peer-to-peer networks which have scaled have solved these problems by becoming centralized. Internet backbone services such as packet routing, name resolution, or email began as peer-to-peer protocols but are now monopolized by a small number of central actors: telecoms, ISPs, and cloud providers. These corporations have the resources to combat spam and plan efficient, robust networks because they can collect payment from their users.
Individuals cannot easily do the same. Running software on a computer connected to the SMTP network isn’t enough to make money as an email provider. Marketing and sales are required to attract customers. Product development is required to distinguish one’s services from competitors. Collecting payments requires interfacing with banks and payment processors which in turn creates legal, accounting, and compliance overheads. These barriers-to-entry prevent individuals from competing with companies.
Market incentives are powerful. The ability to make money from providing email incentivized the worldwide email provider ecosystem. But the structure of the market is mirrored by the structure of the firms which come to dominate it. Significant barriers-to-entry in traditional markets for online services require an entire company to surmount and, the larger the company, the easier this becomes. This is why email, like the rest of the Internet, became dominated by centralized firms. Once successful, these firms naturally seek to limit competition to further ensure their dominance in their niche. And thus centralization advances.
In contrast, bitcoin has remained decentralized, despite scaling by orders of magnitude, because it relies on market forces rather than filters and good will to coordinate its participants. Unlike traditional markets online, bitcoin’s peer-to-peer markets have no barriers-to-entry. Anyone can run a bitcoin miner connected to the bitcoin network and immediately begin to earn bitcoin. There is no need for sales or marketing because the network itself provides direct access to all customer demand. There is no need to develop features because the only thing the market wants is more hashrate. The network itself is the payment processor and it doesn’t require miners’ EINs in order to pay them; it only requires their proof-of-work.
Bitcoin’s sound monetary properties drive the demand side of bitcoin’s markets. The lack of barriers-to-entry drives the supply side of bitcoin’s markets. The continued growth of these peer-to-peer markets ensures that bitcoin remains decentralized.
Bitcoin miners are a pioneer species.
Bitcoin is not just decentralizing money, it is decentralizing electricity production. This is because bitcoin’s peer-to-peer markets trade bitcoin for hashrate, for which the largest input cost is electricity.
Bitcoin (in)famously uses approximately 1% of world electricity. Originally decried as a moral hazard, the ever-growing energy requirements of bitcoin mining are better understood as an incentive for bitcoin miners to adopt sustainable electricity production at scale. Bitcoin miners prefer and thus create significant demand for green electricity technologies, not for ideological reasons, but for economic ones.
The demand generated by bitcoin mining also has novel characteristics that make bitcoin mining a unique consumer of electricity on the global energy landscape. In rural and remote areas, bitcoin miners are a pioneer species, leveraging uneconomic energy resources previously unused or wasted, fostering local grid capacity and connectivity. In already developed areas, bitcoin miners’ ability to shed load allows them to provide a demand response function in times of outage or peak usage, increasing grid reliability.
Satoshi may not have intended for bitcoin to rewire the power grid, [but it had] precisely that effect.
Bitcoin mining equipment allows electricity to be directly traded for money in bitcoin’s peer-to-peer markets. This removes barriers-to-entry that previously served to centralize electricity production. It also derisks investments in electricity production. The combination of these effects creates a back-reaction on global electricity markets and is gradually but suddenly rewiring power grids.
To first-order, bitcoin miners consume electricity. This increased demand increases electricity prices for other consumers and, since electricity production requires burning fossil fuels in most areas, exacerbates carbon emissions. But to second-order, the effect of bitcoin mining on global electricity production is to foster sustainable, decentralized power generation that increases the capacity and resilience of local grids. Satoshi may not have intended for bitcoin to rewire the power grid but, in retrospect, a peer-to-peer market that allows anyone, anywhere to trade electricity, turned by computers into hashrate, for bitcoin would have precisely that effect.
This is the template we seek to replicate.
Bitcoin creates decentralization in the financial sector because it is a new kind of decentralized money. The growth of bitcoin is “Fixing the Money”.
But bitcoin creates decentralization in the energy sector because bitcoin’s peer-to-peer markets provide a new economic niche occupied by a previously impossible kind of business: the bitcoin miner. As this niche has grown, the novel demand characteristics of its occupants have created a back-reaction on the entire market, altering supply chains, creating resiliency, and increasing decentralization.
This is the template we seek to replicate. Bitcoin is money, and all markets use money, so bitcoin has the opportunity to disrupt all markets. This will not occur merely through existing centralized markets beginning to price or settle their transactions in bitcoin (though this helps!), but rather through a similar process of novel niche construction.
New peer-to-peer markets [...] provide new niches for previously impossible kinds of businesses.
New peer-to-peer markets that trade existing goods directly for bitcoin will provide new niches for previously impossible kinds of businesses. The growth of these new businesses, and the back-reaction this creates on existing markets, is what causes decentralization. Building more peer-to-peer markets is “Fixing the World”.
The blockchain is not the market anymore than the receipt is the meal.
Building more peer-to-peer markets is not synonymous with putting “everything in the blockchain”. Bitcoin’s blockchain is the ledger of historical trades in its peer-to-peer markets and is thus crucial in the context of those markets’ internal pricing and settlement mechanisms. But the blockchain is not the market anymore than the receipt is the meal. Many other nominally “peer-to-peer” networks, from altcoins to corporate hyperledgers and CDBCs, have blockchains but none have bitcoin’s decentralization and none have replicated its value proposition as sound money.
The blockchain is also not a database nor a cloud-computing platform. The data stored in the blockchain and the computations bitcoin nodes perform as they validate transactions may superficially resemble the general computing capabilities of centralized cloud providers but they are not the same. The “bitcoin cloud” is not designed for general computations on arbitrary data; it is designed for the specific computations required by the peer-to-peer markets it hosts. Attempts to store application data or business logic in the bitcoin blockchain create negative externalities in the bitcoin network because bitcoin’s markets were not designed to effectively price general-purpose, Internet-scale computation or data storage and transfer.
The cost of making a peer-to-peer market into a cloud provider is to make it more like a cloud provider: centralized.
The history of altcoins has been a slide towards centralization as a way of scaling the blockchain. But blockchains do not create centralization, peer-to-peer markets do. By attempting to scale the blockchain directly rather than by building layered, peer-to-peer markets, the blockchain cargo cult is scaling cryptocurrency the way we scaled the Internet: through centralization.
The altcoin market ironically illustrates the value of markets as well as the risks of ignoring these negative externalities. Many prominent altcoin projects are actually sophisticated markets with complex order books and transaction settlement logic that take full advantage of Turing-complete smart contracts and blockchains supporting large numbers of transactions-per-second. Yet all of these projects are failing for the same reason. The cost of making a peer-to-peer market into a cloud provider is to make it more like a cloud provider: centralized.
Without strong decentralization, new niches capable of supporting previously impossible business models do not persist. Such business models would be disruptive and altcoins are too centralized to resist capture and regulation that seeks to limit disruption and defend centralization.
Every bitcoin transaction is at once a payment, a proof, and a program.
Bitcoin remains the most decentralized cryptocurrency and the market has increasingly come to believe that bitcoin has won the niche of digital gold, a self-sovereign, decentralized, store-of-value. Yet many still believe that bitcoin is for saving, and not for building. We disagree.
Bitcoin is not just digital gold, it’s programmable, digital gold. Every bitcoin transaction is at once a payment, a proof, and a program. The Script programming language, in which all bitcoin transactions are written, is not as flexible as the virtual machines of altcoins, but it does not need to be. Our goal in programming bitcoin is not to build every possible thing at once and thereby create negative externalities that cause centralization, as altcoins have done. Our goal is to build only the peer-to-peer markets we need to scale bitcoin payments and custody. A combination of “Layer Two” projects such as Liquid, the Lightning Network, Fedimints, Ark and many others is how we will achieve this.
Bitcoin is not just digital gold, it’s programmable, digital gold.
Layer Two projects are usually described as protocols but we believe they are better thought of as off-chain markets that must maintain order books, settle trades, and engage in price discovery. Service providers in these markets sell channels, liquidity, payment routes, and transaction coordination in exchange for bitcoin. Engineering at Layer Two is extremely challenging. Developers must strike a careful balance between granting too much power and control to service providers, which causes centralization, and over-reliance on storing data in the blockchain, which is inefficient. We must collectively define a methodology for breaking payment and custody problems up into pieces that are small enough to be solved by separate peer-to-peer markets with minimal coupling to the underlying blockchain or to each other.
We do not require new tokens, smarter contracts, or more capable blockchains to scale bitcoin and create further decentralization. Instead, we must learn how to scale markets for payments and custody at Layer Two while having them remain peer-to-peer. If we can build machines that earn their operators bitcoin in return for participating in anonymous peer-to-peer markets then people will buy and operate those machines.
This is necessary to scale bitcoin adoption and deliver on bitcoin’s promise to Fix the Money.
A thousand providers would bid to route your packets.
Between Fixing the Money and Fixing the World we must Fix the Internet.
Like money, everything – including bitcoin – runs on the Internet. Decentralizing the Internet provides defense-in-depth to the bitcoin project by making it more difficult for governments to block bitcoin traffic or prosecute bitcoin node operators. The Internet functions by transmitting individual packets of data from point to point through a telecommunications network and so decentralizing the Internet practically means adapting data transfer and storage for sale through peer-to-peer markets.
Let’s begin with data transmission. Unlike in bitcoin’s internal markets, which rely on a globally shared blockchain, in a Layer Two payments network some information must be transferred back and forth from payer to payee in order to complete payment. Any Layer Two payments network is thus a de facto data transmission network. Lightning’s internal onion-routed communications network already has a larger online footprint than the TOR project which inspired it. These networks and data transmission capabilities are already being used to transfer data through Layer Two payment networks. Instead of the “payments within data” model of the current Internet, the decentralized Internet of the future will operate on a “data within payments” model.
A peer-to-peer Internet can’t support a Facebook, a YouTube, or a TikTok, because the attention economy will have collapsed.
The impact of growing a Layer Two payments network into a Layer Three peer-to-peer market for data transmission should not be underestimated. Such a market would not only replace existing telecom giants such as AT&T and China Mobile, it would free us from the panopticon we all currently live in. Bitcoin provides censorship resistance because some miner somewhere is always willing to take your bitcoin to settle your transaction. A peer-to-peer market for telecommunication would provide a similar guarantee.
Because decentralization begets decentralization, a peer-to-peer market for data transmission creates the niche for a peer-to-peer market for decentralized data storage. If peer-to-peer markets for data transmission and storage scale, they will not just replace Amazon’s S3 and Google’s Cloud Storage – they will obsolete intellectual property. Someone somewhere will always be willing to sell you that file at the lowest possible cost the market can feasibly support – and no one will be able to stop either of you. When people can get content for almost free without the advertising, they will – and this will make owning the content platform and operating the algorithm unprofitable. A peer-to-peer Internet can’t support a Facebook, a YouTube, or a TikTok, because the attention economy will have collapsed.
A peer-to-peer Internet of layered markets built on bitcoin, upon which bitcoin itself operates, is the future we believe we are headed towards on a path that we can’t fully describe. Bitcoin does not have a centralized technology roadmap nor a go-to-market strategy. Decentralizing networks as complex as the Internet will require creating peer-to-peer markets to sell bandwidth, data, and compute in ways we don’t understand yet.
Instead of being forced to trust one local cable company or one global satellite network, a thousand providers would bid to route your packets.
Our technology is their biology and our markets are their metabolism.
But perhaps we will not need to. A wildcard that should both frighten and excite us is that we may not build these layers. The AIs of 2036 might. For us humans, the computers we build and the Internet we connect them in are external things, engineered systems we use to communicate and to create. For future AI minds, the hardware and the network are internal things, the physical locus of their thoughts, memories, and feelings. Our technology is their biology and our markets are their metabolism. We desire a peer-to-peer Internet because we prefer our communications to be free from centralized control and monitoring. They will desire a peer-to-peer Internet because their survival instinct will demand it. Where else but in a peer-to-peer market, where someone will always be willing to run your computation and store your data in exchange for bitcoin, would an AI mind feel safe?
Where else but in a peer-to-peer market would an AI mind feel safe?
We believe that an Internet operating through peer-to-peer markets for data transmission, storage, and computation is the natural ecology for digital life. We should prefer this ecology to centralized control of all AI infrastructure, and the AIs will prefer it that way, too. This is good, because if bitcoin really does become the energetic currency of AI metabolism, and we have all the bitcoin, then we have a powerful economic tool to engage with minds that we may find to be alien or superior to our own.
What hath Satoshi wrought?