Harnessing The Power Of Prediction To Scale Lightning

The Lightning Network is a powerful Bitcoin scaling solution, offering increased transaction throughput and lower fees while maintaining the safety and security assumptions of the underlying blockchain. However, the unique topology of the Lightning Network, with liquidity locked upfront in pairwise channels, can introduce friction in the movement of value. As liquidity becomes more fragmented, it is less likely to be available for payments, ultimately slowing down the velocity of money on the network and making it more expensive. 

From the outset, it was apparent to the Lightspark team that scaling Lightning would hinge on capital efficiency — ensuring that value can move quickly on the network without requiring capital to be idle. Capital is expensive, so scaling Lightning means ensuring that liquidity is where it is needed, when it is needed.

This article briefly explores how better measurement, applied economics, and AI and ML methods can drastically improve capital efficiency and thus how participants experience Lightning. By rethinking how you connect to the network and by leveraging AI for prediction, it is possible to significantly improve capital efficiency and payments performance.

Better Measurement

"If you can't measure it, you can't improve it," as Peter Drucker supposedly said. In the context of the Lightning Network, many aspects were not well measured. Before we could optimize the network's efficiency, we first needed to go back to basics and develop better measurement. 

Existing Lightning services, operators, and websites that rank participants on the network place a lot of emphasis on popularity, often relying on simple heuristics. While this may seem like a sensible approach, it often falls short from an economic perspective. When deploying capital on Lightning, it is essential to allocate it based on what you are trying to achieve. This means that your configuration should heavily vary depending on your business needs and your exact composition of sent, received, or routed transactions.

Maximizing Conductivity

To build a customized approach for Lightspark clients, we developed a new metric, which we named "conductivity." Drawing inspiration from the world of physics, where conductivity refers to a material's ability to conduct electric current, Lightspark’s conductivity metric captures a Lightning node's capacity to effectively move value through the network. By ensuring that each node on the network has high conductivity for its business purpose, we can ensure that everyone gets the most out of the capital they bring to Lightning. This also means avoiding competition with well-served existing routes, and only establishing connections that have a positive return to the capital dedicated to them. Focusing on a customized conductivity metric allows participants to bridge otherwise disconnected parts of the network, and not crowd existing routes. This approach is similar to how airlines scout opportunities to serve underserved routes and regional airports.

Dynamic Adjustments

AI holds significant promise for the Lightning Network: it can evaluate network segments that lack liquidity or suffer from an overabundance of it ahead of time. It can identify reliable routes for successful payment, iteratively improving participants’ conductivity as models evolve. As payments shift liquidity repeatedly in the same direction — for example, because consumers pay merchants on the network — this approach can also help rebalance liquidity and prepare for more transactions. When the state of the network changes, AI can be used to continually refine conductivity metrics and liquidity for optimal efficiency. At scale, this ensures that liquidity is where it is needed, only when it is needed, making the network more capital efficient than existing payment rails. 

The combination of advanced metrics, economics, and AI enables the Lightning Network to scale. By rethinking how participants connect to the rest of the network, and continuously adjusting liquidity based on business needs, we can make Lightning truly efficient. As Lightning becomes more capital efficient, it can support greater payment volumes at a lower cost, larger transactions, and a more thriving and innovative ecosystem. The confluence of AI and Lightning promises to usher in a new era of digital payments.