The author is interested in understanding the risks associated with these parties and why routing through them would be trustless. Additionally, they are curious about how users of the network would find and select intermediaries. While building trust-based level 2 protocols is a good idea, it may not work without explicit trust. Opening a channel requires trusting the counterparty up to a certain amount, and if they disappear, the risk is capped at that amount. In contrast, the existing banking and monetary system allows for minimizing counterparty risk by shifting unwanted exposures. However, the problem lies in systematic risk, such as a failure of the entire banking system. If counterparties are interconnected, failures can propagate in unexpected ways. For example, A might trust B to route or clear and not trust C, but B might have exposure to C, making it impossible to diversify A's exposure.