DLT Isn't Disintermediation
Written by Topper Bowers
"So we’re in this very interesting period now where we’ve got this major new Lego block, which is this idea that you can build trusted things, and then this Lego block . .. and it’s the best Lego block, it’s awesome, that’s why I’m so excited. So it just totally changes what you can build with Lego blocks."
- Chris Dixon
I've recently been exploring the idea that DLT isn't about disintermediation at all. My ideas aren't fully baked, but I want to explore by writing.
A lot of talk in the DLT space is about getting rid of trusted intermediaries. Sometimes: "kill the banks!" comes up. I've always had a visceral distrust of these kinds of statements. Part of the reason I started Tupelo was to allow a path forward that doesn't tie itself to maximalist thinking.
I'm starting to understand DLT more and more as the infrastructure of trust. This line of thinking makes the applications of DLT boundless. Thinking about DLT in this way also makes it more palatable for the status quo.
Let me explain.
There are a whole variety of companies where the major cost of what they are doing is providing trust.
I would put Lyft in this bucket. There are other groups providing trust in a real-world sense that don't care what the infrastructure is. I would put banks in this bucket.
Take the Lyft example. The main value they provide is reputation management and pricing models. The rest is cost.
Lyft certifies that drivers are safe and licensed. Lyft certifies that passengers will pay their bills and suffer consequences for bad actions. These certifications are immensely important. I would say this is actually the core of their business and what allows the marketplace to flourish.
Another key service is pricing model. Lyft attempts to optimally distribute supply and demand through pricing.
In order for Lyft's market to work, the users need to trust their counter-party and also trust the calculated price. More specifically, the rider's and driver's software needs to trust these things.
Before DLT, the easiest way to accomplish this trust was to setup a central server and route all requests through this trusted intermediary. This is expensive. Lyft's S1 filing shows that it spends north of $100mm/yr at AWS.
What a DLT allows is for trust to permeate the entire marketplace that Lyft has setup. DLT allows drivers and riders to communicate directly with each other while still allowing trust to flow from Lyft.
The best part about this scenario is that we're not talking about getting rid of Lyft. Lyft's core business is extremely valuable.
We *are* talking about saving the company some big portion of $100mm a year. We are also talking about more resiliency and reliability for Lyft's customers as well.
This post is getting long, so I'm going to save the technical details of this for another post (we'll get to that tomorrow) but I'll end this with; DLT makes simple rules easy to implement at the edges. Things like "this driver is only in one place at a time" or "this rider canceled their ride, so needs to pay $5" can be handled at the edges of the network – without talking to Lyft's network.
DLT allows systems to self organize, distribute trust and act together. It doesn't need to disintermediate anyone. Instead, it can empower. It can make systems efficient.
DLT is the digital infrastructure of trust and most of what we do builds on some amount of trust.