Recently, I helped someone troubleshoot an IBC cross-chain issue that got stuck, and I took the opportunity to revisit the question of "who to trust." Frankly, a single cross-chain transaction isn't just about clicking a button: you have to trust that both chains won't arbitrarily change their states, trust that their light clients/validation logic won't crash, trust that relayers are willing to forward messages (or at least that someone will), and also trust that the channel/timeout settings are correctly configured, or else the message just "sleeps on the way." Bridges are even more straightforward: adding more signatures/multisig/operations means adding more human and compliance pressures. Recently, the community has been arguing about privacy coins, mixing, and compliance boundaries, but I think cross-chain is more practical—once certain components are targeted, the first thing to change isn't the technology, but "who still dares to help you transfer."

I'm not focusing on big goals right now; I see cross-chain as a two-step process: first, small amounts for testing and confirmation, then increasing positions; sorting out small pitfalls like transaction order/nonce first, to avoid blaming the chain, bridge, or the world in the end... Anyway, details are king.