Digital Letter 23

Remember Non-Fungible Tokens or NFTs for short? They were a part of the big cryptocurrency boom that Bitcoin started, but they never really went anywhere, and wound up with a reputation for being mostly a vehicle for various frauds (a number of which were really just poor business decisions). The big problem with NFTs was the lack of a compelling use case. While some ideas have been floated, none of them really solved an identified problem better than what was currently available. And that, for the most part hasn't changed.

But it might. I was listen to a recent episode of the Decoder podcast, where host Nilay Patel was speaking with Lyft CEO David Risher. One of the things that they talked about was what Lyft was doing, and could do, for drivers, and Mr. Risher floated the idea of Lyft giving out what would basically be letters of recommendation. The letters would generally outline why Lyft felt they were an excellent driver, and recommend them for other service roles. They would also be created by generative automation.

And that's where NFT technology might prove useful. With a standard reference letter, it's always possible to contact the person who wrote it, and authenticate it that way. If a computer has written the letter, that method doesn't work. It's possible to keep a database of all of the letters, but then the database would need to have enough information to positively identify the recipient; one doesn't want all the John Smiths to be able to claim the achievements of one. And that would make the database a target for information thieves.

Given that an NFT is basically just a digital file of some sort with a blockchain-enabled certificate of authenticity attached to it, it could solve the problem. Whomever the driver shares the letter with can verify that it was actually issued by Lyft, without exposing some sort of database of personal information.

To be sure, I'm not into the technology of cryptocurrencies, and so this may be over-engineering something that has a much simpler solution. But it strikes me, at least at first thought, as being a potential use case for a technology that's been searching for one for some time now.