DARPA is known for seeing futures first, and then making them real. The Pentagon’s blue sky projects agency laid the groundwork for everything from the internet to driverless cars, and continues to explore the edges of technologies such as battlefield autonomy and paradigm-shifting cheap space launches. All of which makes a recent DARPA request for information on blockchain almost an indictment of the technology itself. If DARPA can’t figure out a responsible, value-generating use for blockchain, who can?
People or organizations with ideas about how the national security community can use blockchain (or, as DARPA puts it, “permissionless distributed consensus protocols”) are invited to respond to the request until Dec. 20, and if selected, may present their ideas at a workshop in February.
So what, exactly, is DARPA looking for? The most common use case for a blockchain is in cryptocurrency, like bitcoin. The cryptographic protocol creates a ledger, which tracks the full record of transactions and makes it extremely difficult to falsify improperly generated inputs. (I find this explainer in comic form helpful). In the coin iteration, these units are then exchanged as a kind of currency, though the variability of price at a moment makes it something more like an unstable commodity than a reliable store of value.
But DARPA is steering clear of currency-like uses. Currency has been a natural outgrowth of blockchain technologies because it provides the clearest incentive for many people to maintain a distributed ledger: keep running the computers doing the math to track cryptocurrency exchanges, and the exchange will periodically reward one ledger maintainer with new units of currency. (Commonly, this is what “mining for bitcoin” means, and it uses enough computing power to have its own trackable environmental impact).
Stay with me.
Rather than wade into the world of cyrptocoins, DARPA is side-stepping it, which means one of the big challenges for any blockchain application they might find value in is encouraging people to use it. The request specifies that “ all means of rewarding participants (e.g., giving them access to computing resources) also constitute a transfer of value; such transfers are within scope of this topic as long as rewards do not consist of money.” So one possible way to distribute the ledger may literally be distributing the ledger, giving people computers to use and maintain on the condition that they keep the ledger program running.
The second topic for DARPA’s blockchain workshop is no less ambitious than the first. DARPA’s looking for methods that can combine economic notions of utility with the computer science world’s understanding of protocol participants as “honest” or “malicious.” If someone has an idea about how people seeking to maximize their benefit from the use of an open ledger can be kept from acting maliciously in that space, DARPA is all ears.
Finally, DARPA is also looking for analysis of the way that distributed protocols tend to centralize, and the vulnerabilities that this centralization can bring. In the case of existing blockchain cryptocurrencies, for example, it is third-party services that manage the cryptocoins of others through password-gated systems that are often hacked and stolen from. Ways to anticipate centralization and mitigate the risks could help keep a blockchain system as secure as the individual pieces in it.
With so much uncertainty, vulnerability, and weirdness baked into the concept, it’s hard to imagine the exact utility DARPA wants to get from blockchain. And that’s partly the purpose of the workshop. Acknowledging that this technology has undergone real refinement and development thanks to blockchains and cryptocurrency, the agency is keeping an open mind about what, if any, function the military can glean from this work.
“These technologies have dramatic implications for the security and resilience of critical data storage and computation tasks, including for the Department of Defense ” read the request. “At the same time, the concrete applications and security of these technologies for the DoD is unclear.”