Lecture 11: The Future of Bitcoin?

What this lecture covers

This is the final regular lecture of the course, and it steps back from Bitcoin's mechanics to ask what else blockchains might decentralize, and whether doing so is actually a good idea. The lecture builds a template for analyzing any decentralization proposal: what is being decentralized, which route is used to integrate with a blockchain, what level of decentralization is achieved, and what security mechanism replaces trust.

Using a running example of a car whose ownership is tied to a Bitcoin transaction, the lecture works through smart property, crowdfunding, paying for a proof, colored coins, side chains, Ethereum-style contracts, escrow disputes, prediction markets, data feeds, decentralized exchanges and Ripple-style IOU networks. It closes with a critical assessment: many of these systems automate what happens when things go right but struggle with what happens when something goes wrong, since disputes and enforcement remain fundamentally human problems.

Key ideas

• Representation: encoding a real-world transaction (such as car ownership) into something a blockchain can express, for example treating a Bitcoin address as a car's authorization key.
• Atomicity: coupling two sides of a transaction, such as a payment and an ownership transfer, into a single blockchain transaction so neither side can happen without the other.
• Routes to blockchain integration: four approaches: directly on the Bitcoin blockchain, embedding extra data into it (colored coins, Mastercoin), a merge-mined side chain, or a separate alternative chain such as Ethereum.
• Levels of decentralization: a spectrum from a single mandatory intermediary, to a threshold of intermediaries, to multiple competing intermediaries, to full disintermediation.
• Security alternatives to atomicity: reputation, escrow with dispute mediation, and trusted hardware, each with limitations when participants are pseudonymous.
• Data feeds: a mechanism for arbiters to assert real-world facts (prices, event outcomes) onto the blockchain, which can be decentralized using a threshold of arbiters.
• Ripple's IOU model: disintermediating currency exchange through chains of trust relationships between friends, rather than through atomicity.
• The human-problem critique: security and dispute resolution rely heavily on law enforcement and courts in the real world; blockchain systems shift almost all the burden onto prevention, which introduces new problems like key-loss and software security.

Walkthrough

Smart property and the car example (2:05)

The lecture introduces smart property through a car whose digital key checks a public key hardcoded at manufacture. The car then watches the blockchain and updates its authorized key whenever the linked Bitcoin transaction is spent to a new address, so a Bitcoin transaction effectively transfers ownership. Combining Bob's payment with Alice's ownership transfer in a single signed transaction, using the same technique as CoinJoin, makes the trade atomic: neither party can walk away with only their half.

Representation, atomicity and disintermediation (13:08)

The lecture names the two technical ingredients behind these schemes, representation and atomicity, and frames the car example as decentralization through disintermediation: the DMV as centralized authority is removed entirely.

Four routes to blockchain integration (15:08)

Working directly on the Bitcoin blockchain is easy to deploy and inherits Bitcoin's mining security, illustrated with crowdfunding transactions that only become spendable once enough inputs are collected, and "pay for a proof" schemes that publish a value in exchange for payment. Embedding (colored coins, Mastercoin) allows richer representations while still using Bitcoin's security, at the cost of limited scripting and blockchain bloat. Side chains merge-mine with Bitcoin and peg value one-to-one but require Bitcoin protocol changes to support. A fully separate chain like Ethereum offers Turing-complete scripts and long-lived contracts paid for with gas, giving maximum flexibility but unproven mining security.

Escrow, dispute mediation and the decentralization spectrum (27:17)

A two-of-three escrow address involving buyer, seller and a chosen mediator handles disputes over a car sale, replacing the court system with a freely chosen mediator. This is decentralization through freedom to choose an intermediary rather than full disintermediation, echoing the earlier lecture on decentralized prediction markets. The lecture lays out the resulting spectrum: single mandatory intermediary, threshold of intermediaries, competing intermediaries, full disintermediation.

A taxonomy of decentralizable things (43:33)

Purely digital things (name mapping, storage, proof of computation, random beacons, lotteries) are straightforward to decentralize. Real-world assets like currencies and stocks can be represented with colored coins, but only work if some entity actually backs the peg to real value. Smart property and trade extend the car example. Complex contracts, including financial derivatives, need reliable data feeds for prices, decentralized here through a threshold of arbiters signing a multi-signature output. Decentralized markets need matching mechanisms too, illustrated with a partial-transaction broadcast scheme and a double-auction variant run by miners. Autonomous agents and Ripple's disintermediated currency exchange, built on chains of IOUs between trusted friends, round out the taxonomy.

Is this actually a good idea? (1:08:51)

The lecture closes by weighing smart property against what it replaces. Real-world security relies on preventive, detective and corrective measures, with law enforcement doing much of the corrective work; a blockchain-based car can only really offer prevention, and adds a new software-security problem of protecting the key itself. Dispute mediation over a "lemon" car sale is shown to be a deeply human process, illustrated by an anecdote about sitting in on a real small-claims hearing, that a fixed escrow mechanism cannot replicate. The lecture argues the real opportunity is not to replace state institutions but to find compelling use cases (such as cell-phone-minutes-as-currency in parts of Africa) and to integrate cryptographic systems with existing legal and regulatory practice rather than against it.

Before you watch

• Review the smart property and CoinJoin-style transaction constructions from earlier lectures, since this lecture reuses them without re-deriving them.
• Recall the decentralized prediction markets and colored coins material (referenced as lecture 9) and the side chains material (lecture 10).
• Be familiar with two-of-three multisignature escrow and zero-knowledge based Zerocoin, both used as recurring examples.

Check your understanding

1. What are representation and atomicity, and why does the lecture treat them as the two key technical ingredients of blockchain-based decentralization?
2. How do the four routes to blockchain integration (direct, embedding, side chains, alternative chain) trade off security, expressiveness and deployment cost?
3. Why does the lecture argue that colored-coin representations of dollars or stock only work if some real-world entity backs the peg?
4. In what sense is choosing your own escrow mediator a form of decentralization, and how does it differ from full disintermediation?
5. Why does the lecture conclude that dispute mediation and security in blockchain-based systems remain fundamentally human problems?