Agentic AI has moved past the hype phase into the aggressive deployment phase. Autonomous systems are now trading, booking, and paying each other with no human in the loop. Animoca Brands has put $10 million behind it, funding developers building on its new Minds AI agent platform. Joining me is the co-founder and executive chairman of Animoca Brands, Yat Siu. Thank you for joining us.
Thank you for having me. It is a great pleasure to be here.
Animoca is already running around 280 AI agents of its own — from scanning code to food price arbitrage. When a developer's agent first makes money on your platform, who gets paid and how does it work?
I now have close to 400 agents myself — around 380 of them. Many are experiments. Think of them as co-workers. For an entrepreneur or someone with many ideas, agents are a godsend. Whenever I want something done, I tell the agent — go do it. Sometimes it fails, but it just does it. Imagine you had at your fingertips a limitless number of employees who could execute on any dream project you wanted. That is how I treat agents. The success rate started at around 50-50 and is now around 80 to 90%. We are heading to a point where you have an idea and the agent comes back with a workable solution that is 80 to 90% there, and you fine-tune it from there. People are making money on Minds — on the skills bazaar, on sales lead generation, on new ideas. Some people have already made hundreds of thousands of US dollars using Minds. And we believe we are getting to a point where you will have billion-dollar startups run by one, two, or three people managing hundreds of agents.
You have argued that agents will turn everyone into a manager. Who gets paid when one person manages a thousand agents doing the work ten employees used to do?
Those ten employees all become managers of their own. Think back in history — people said we would not need secretaries because of computers and automation. Turns out we have more demand for employees than ever before, just different ones. When you have a net gain in productivity, you get an explosion of opportunity and creativity. Much of the work 50 years ago was menial and labour-driven. Today it is creative. Those ten employees who fear they are out of a job — they are going to be doing much more creative work because the agents will release them. The skills we need are going to be focused on creativity, divergent thinking, and management, as opposed to executing a particular task. Subject matter expertise matters even more because now you can direct agents more specifically. Good managers who know their domain — they are going to thrive. Bad managers who do not know anything — they are going to be out of a job. But that has always been true.
You have said the $900 billion online ad market is going to shift entirely into a transaction economy. How does paying to be an agent's default choice differ from the ad system it is replacing?
When we optimised for SEO, we wanted traffic from Google searches. Now with Gemini, when was the last time you clicked through to a website that the AI summary pointed to? Some websites are reporting 80% drops in traffic. So advertising revenue based on clicks is collapsing. What happens is you get paid for the content instead. That $900 billion — close to $2 trillion of advertising revenue — is going to shift to a transaction-based economy. The end user pays $20 a month and a fraction of that goes to the content provider instead of to advertising platforms. The key difference is that in a centralised platform like Facebook, the algorithm acts in Facebook's interest — more clicks, more anger, more harm. Facebook just had an $18 billion settlement related to harm to children. Why? Because the platform only cares about clicks and attention. The agent is different. The agent belongs to you. It acts in your interest — what is healthier for your mind, what is better for your productivity, what is good for your wellbeing. That is a fundamental difference. If we are all running agents that act in our interest, I think that is a very positive shift.
Thank you so much for joining us.
Absolutely. Thank you for having me.
