Anthony calls Verona “anti-AI” even though the team is building for AI agents and he already uses an agent to book flights. His @CryptoCoinShow interview cleared up quite a few things for me. @BurntBanksy (Anthony Anzalone) is pretty skeptical of promises that everyone will soon hand their online shopping over to AI. He has doubts about next year, too. Most people are still getting used to a chatbot as an alternative to a search engine. Trusting it with money and letting it act on its own takes a little longer. Meanwhile, Anthony already has some fairly ordinary examples from his own life. He kept putting off sorting out his airline miles. Eventually, he asked an agent to check his email, find his previous flights and sort out the mileage claims. According to him, it worked: miles from flights he had taken three to six months earlier appeared in his American Airlines account. At the end of the full interview, he also mentions using an agent to book his flight back from Singapore. These are easy tasks to relate to. Someone had something sitting on their to-do list for months. Now they can hand it over. @verona_dev is being built with that gradual adoption in mind. Anthony uses the hockey analogy: a startup needs to head where the puck is going. Waiting until millions of people use agents every day before building the infrastructure would leave them starting pretty late. His answer about MCP and x402 particularly caught my attention. Asked about connecting agents to Verona, he says that work is already done. According to him, x402 has been live for quite some time, and the integrations are helping projects and making payouts easier. The next job is to build applications that need those capabilities. The team’s main focus remains verifying data and storing those proofs so they can be reused. That’s how I understand his “anti-AI” comment. As the internet fills with generated content, being able to verify a particular fact becomes more useful. An agent will eventually have to act on that information, sometimes with money involved. The conversation also makes verUSD easier to understand. Anthony says outright that he doesn’t need to convince people to hold the stablecoin. He wants payments flowing through the system. A user completes a mission, gets paid, transfers the money to a bank account or spends it with a card. They might have no interest in the name of the currency used along the way. He gives payouts for Starbucks missions as a specific example of payments that will go through verUSD. Having their own stablecoin also gives the team more flexibility in how they set up the economics. Anthony explains that working with USDC meant operating under someone else’s rules, including restrictions on offering yield to projects. The $100 million figure is worth unpacking, too. In the interview, he describes roughly $60 million in committed revenue tied to future payments, plus another $40 million in institutional capital. It would be premature to count that entire amount as payment volume that has already gone through the system. A few other details are easy to miss among all the AI discussion. The team is discussing proof of reserves with a partner. Anthony sees a role for ZK here: proving the relevant information while preserving confidentiality. At the time of recording, a post-quantum upgrade was being prepared to improve the blockchain’s resistance to quantum attacks. He expected it to launch around the time the interview came out. What interests me most is how these different efforts are starting to connect. Products already need verification. Payments flow through those products. And agents will be able to use the existing connections as people start handing them more tasks. ━━━━━━━━━━━━━━━━━━━━━━ So🙄now I’m much more interested in watching which applications use these capabilities and which payments actually go through verUSD. The story about the forgotten miles is a good example of where it can begin: someone finally gets a task done that they’ve been putting off for half a year.