AI Agents as Everyday Helpers
Yat Siu from Animoca Brands thinks AI agents could change how people interact with Web3 technology. He uses nine different AI agents in his daily life. One handles family travel planning across time zones. Another manages investor relations and contact lists.
What makes these tools interesting is how they learn. Siu shared an example where his flight from Tokyo was delayed. The agent initially checked only one website. He told it to check five sites using APIs. Within minutes, the agent built a flight tracking skill that notified his wife about his arrival and arranged pickup.
“The real value isn’t me talking to the AI,” Siu said, “but the AI talking to many people in a group.”
Making Technology Accessible
Most people use AI like a better search engine. They ask questions and get answers. But true AI agents act independently. They push information to you through familiar channels like Telegram, email, or soon WhatsApp. This feels more natural, like human conversation.
Animoca launched Animoca Minds specifically for regular users, not just technical experts. Siu compares it to early web servers. Setting up Apache used to be difficult and risky. Now services like AWS handle it with one click.
Other tools exist, but they’re often complex. Some require access to all your emails and files, which raises privacy concerns. Animoca Minds uses chat interfaces that focus on calendars, flights, and workflows without deep personal data access.
Current Web3 Challenges
Web3 currently has about 50-60 million users, but growth has stalled. Major platforms like Apple, Google, and Steam limit or block NFT functionality. AI agents could change this dynamic.
Today, AI summaries keep people on Google while websites lose traffic. Soon, agents might fetch information directly and push it to users without website visits. This shifts power back to users while keeping payments and incentives on blockchain rails.
Wallets might not be the main attraction for mass adoption. Most people won’t join Web3 for trading, but for fun, help, or games. The wallet becomes just another feature. When you want to play a game requiring an NFT, the agent handles the purchase, bridging, and market interactions automatically.
Gaming’s Role in Web3
Some critics say gaming failed in Web3, with finance being the only successful application. Siu disagrees. Gaming expands what’s possible in this new internet environment.
Finance will always be larger than gaming in terms of market size. JP Morgan deals with $800 billion while EA sold for $60 billion. But early Web3 games with $20-50 million valuations were solid foundations. Animoca entered Sandbox at $4.5 million and Axie Infinity at $8 million, both now worth hundreds of millions.
GameFi essentially gamifies finance. Platforms like Polymarket function like betting games. Robinhood’s leaderboards appeal to younger generations who enjoy competitive elements. Gaming tests these concepts in accessible ways.
Take Anichess, a chess game with a CHECK token. It has a $60 million fully diluted value, not billions, but works well for chess enthusiasts. Backed by Chess.com and Magnus Carlsen, it maintains good liquidity and holds value.
The lesson seems to be starting modestly at $10-20 million and growing steadily. Avoiding greedy pumps leads to sustainability at $30-40 million valuations that keep fans happy without crashes.
The Path Forward
Siu sees AI agents as crucial for mass adoption. They could onboard billions of users by handling tedious tasks and building persistent memory systems. Blockchain ensures ownership and fair payments through decentralized infrastructure.
Animoca remains optimistic about gaming’s future in Web3. Combined with AI agents, blockchain technology could become ubiquitous in the background of everyday applications.
Perhaps the future isn’t about convincing people to use blockchain directly, but about making technology so seamless they don’t notice it’s there. The friction disappears, and what remains is simply useful functionality that happens to run on decentralized systems.
![]()



