Rob Desideri

I've spent most of my career in the arteries of global commerce: derivatives, risk management, and institutional product and systems design across equity and debt markets, with detours into commodities and energy. Somewhere along the way, before "fintech" became a category with its own conference circuit, my attention drifted up the stack, from the mechanics of markets to the harder problem of how people and institutions coordinate at all.

Every institution and governance architecture has a scale and a speed it's honest at, including state-shaped ones. To different extents, each is good at solving small-group problems: shared context, legible intent, people who see each other often enough to build trust. The failure mode is less often an architecture defect and more often people trusting the system past the point its assumptions still hold, whether that's a trading desk, a company, or nations trying to trade with each other without settling everything through one dominant currency. Capital gets trapped in the gaps, and humanity pays for it in quiet and not-so-quiet workarounds, spreads, and dead time.

As more sovereign currencies and their digital proxies compete for a role in cross border settlement, algorithms are already setting the pace: markets move faster than the institutions meant to govern them, which is its own kind of double jeopardy, speed without oversight on one side, oversight without speed on the other. Whatever equilibrium emerges will still be decided by people and institutions, not by protocols alone. Better tooling won't replace that diplomacy. It's what would let diplomacy happen at the speed markets already move.

What I keep coming back to is something like quick coherence: can a group actually create and hold a shared understanding as it grows, or does it just look that way from the outside? Can we tell the difference between real agreement and cheap talk, the kind that costs nothing to offer? I don't think we've built that tooling yet, and I think it's some of the most useful work available right now.

There are no perfect answers. What's required is thoughtful individuals willing to bring new lenses to the mess, using whatever tools the machine's newly handed us. For me, that also means copious quantities of coffee alongside the compute.