What are Blockchains and Distributed Apps really good for?
A lot, but not as much as some ICO enthusiasts claim…
Lightspeed is a huge advocate for distributed ledgers and crypto, having invested in this space since 2013. But we think that some of the euphoria that “blockchains change everything” is fundamentally misplaced. What blockchains change, they will change a great deal, and for the better. But blockchains will not change everything.
We are all seeing more and more white papers for ICOs touting various benefits of applying decentralization, blockchains, etc., to various markets. Since you can now get an ICO white paper written starting from $30 on Fivver, it is not surprising that some of these claims are not well thought through.
Here is my take on some of the purported benefits:
Speed and Efficiency: Typically not true, but depends on what you’re comparing it to. If you’re comparing it to an existing legacy system, that may be true. If you’re comparing it to a de novo solution built with current tech, a decentralized system involving miners of any sort can never be as fast as a single modern database.
Scalability: Typically not true. There is typically a scaling penalty with distributed systems, not a benefit.
No Single Point of Failure: Typically true. But for most big companies, not a big deal. If the problem you’re going after is big enough, and the companies currently solving it are big enough, then they’ve built a ton of redundancy into their current solution. Now even with a ton of redundancy, AWS (as an example) goes down from time to time, but not for very long. So this is a benefit but it’s a 3–5 nines level of benefit.
Censorship Resistance: Typically true. There are two potential sources of censorship:
- The first source of censorship is a market participant. Examples of this include AirBnB which banned some white supremacists from renting in Charlottesville*, Cloudflare declining to protect the Daily Stormer from DDoS attacks, and Uber and Lyft banning Laura Loomer after her calls for a Muslim-free ride sharing company and other anti-Muslim tweets following a terrorist attack in Lower Manhattan. (*Obviously I’m not trying to support hate speech. But I think it is actually an important thing to note that the only examples of company censorship have been to censor hateful behavior. In the absence of the examples, this is a pretty theoretical conversation. I’m not aware of other sorts of censorship, but would welcome other examples.)
- The second source of censorship is a government. Examples include China’s great firewall, Austin’s ban on Uber, and San Francisco’s registration requirement for AirBnB listings. A distributed ledger system allows market participants to bypass government regulation. However, because it is public and immutable, it also makes the prosecution of those violating regulations easier to prove. This was ultimately the downfall of Ross Ulbricht of Silk Road. Once a single link between his bitcoin accounts and his real identity was found, the blockchain provided evidence of many related transactions.
Enable Trustless Transactions: Typically true, because of the immutability of transaction history for each participant. But in many instances not a big deal. If the use case is for frequent transactions between strangers, this is a benefit. Money is a great example. Knowing that a check is good or a credit card has money behind it is important for e-commerce.
But there are many instances where trust is solvable in other ways. One example is when the transactions are between a smaller group of market participants who transact with each other frequently, e.g. the settlements of stock trades. In most instances, the big investment banks know who they can and can’t trust, and how much they can trust each other. Just before Lehman went under, other banks stopped trading with them because they knew that they could no longer feel confident in the counter-party risk.
Another example is when trust can be built through historical reviews, monitored and maintained by a single trusted party. AirBnB reviews are trusted by travelers, even if the individual property owner is not. This means that transactions between untrusted parties can still occur.
Enable Network Effects: Typically true. But mostly useful only for two-sided marketplaces and other businesses with network effects. Putting coins in the hands of one side of a marketplace (typically the supply side) elegantly solves the chicken and egg problem of a two-sided marketplace. Bitcoin with its money use case is an example of a single sided marketplace with network effects. But there are many businesses where there are no network effects and this benefit does not apply.
So where does this leave us? I’m excited about DApps that are focused on any of the following key advantages, with more being better:
- Avoiding single points of failure where 5 9s isn’t enough.
- Enabling censorship resistance where there is a core set of users for whom this is the MOST important feature, one that they are willing to trade performance and price for.
- Enabling trustless transactions where transactions between strangers is common and reviews are insufficient.
- Kickstarting network effects.
*Image source: t3n.