AI Agents Do More Than Just Place Orders: The Next Entrepreneurial Opportunity Lies in Matching and Trust

Breaking down how AI agents can expand Uber-like matching to complex transactions, highlighting that bilateral networks and trustworthy transaction mechanisms are the real moat.

Written by: Guy Wuollet

Compiled by: AididiaoJP, Foresight News

Uber simplifies demand into one thing: calling a car from A to B. The network is responsible for finding drivers who are willing to take the order at a suitable price. The more drivers there are, the easier it is for passengers to get a ride; the more passengers there are, the more drivers are willing to join.

The author believes that AI agents can expand this matching network to more goods and services. Agents can articulate complex preferences on behalf of individuals and then find suppliers willing to take the order. Some transactions exist today but are too expensive or convoluted for the average person to participate; some transactions are too troublesome for anyone to organize; and there are some demands that people haven't thought to raise yet. For entrepreneurs, these are all opportunities.

Can Uber-like Networks Be Used to Arrange a Meal?

Organizing a dinner with friends is a familiar challenge. Everyone has different schedules, budgets, and tastes. Often, one person is responsible for collecting preferences in a group chat and searching for restaurants; if the options are unsuitable, they have to communicate again in the group. On the restaurant side, there may be empty tables available on Tuesday nights, but it's hard to reach a group that hasn't finalized their plans.

An agent can take over this task. Your agent coordinates with friends or their agents to form a request: six people, in the city, with a budget of no more than $60 per person, available between 5:30 and 8:00 PM, and willing to eat anything except sushi. The restaurant agent quotes based on the availability of tables. One restaurant might say that the price is better for a 5:30 PM seating; another might offer a family-style menu for the group. Your agent matches the quotes with everyone's preferences and places the order within the agreed range.

If a restaurant has an available table at six o'clock, it can directly quote the group willing to eat earlier, even before they have chosen a restaurant. The group has preferences and budgets; the restaurant has capacity and acceptable prices. What’s missing is a network that allows both parties’ agents to find each other and agree on a purchase.

Uber has proven that a network can bring buyers and sellers together around relatively simple demands. Agents can extend this model to more complex purchases, such as a group dinner. Most networks are built for people; now, they need to be built for machines. Entrepreneurs can create such markets: agents bringing in requests while suppliers compete to fulfill them.

A 'Machine-Use' Market is Needed Between Agents

The crypto space already has models for organizing such requests: intents, or intention orders. Users simply state the desired outcome, and suppliers compete to deliver. For example, with UniswapX, users sign an order specifying what token they want to swap and what conditions they can accept; suppliers compete to match from different sources of liquidity.

Agents can extend the intent model to everyday commerce. Large language models can turn a conversation into requests that suppliers can execute, and agents can then search for quotes and negotiate details. Arranging a small purchase, like a dinner for six, could be cost-effective enough to justify doing so.

Uber’s usefulness relies on network effects. More complex purchases can also create similar network effects. The more restaurants that join, the more matching opportunities for diners, especially when the group has specific requirements. The more groups there are, the more opportunities restaurants have to fill empty tables, making it worthwhile to integrate a reservation system and provide more tailored quotes. The more participants in a city, the more types of requests the network can accommodate.

Entrepreneurs can start with dining in a city and then expand outward. The initial product must save users enough time or money for them to invite friends to use it; restaurants need to see reservations they couldn't originally get. Only when both sides achieve results will they return and invite others.

Ensuring Both Sides Benefit

The author believes that consumer-facing agents can be used for free, charging through a model similar to order flow commissions. Users specify what they want and how much they are willing to pay, while suppliers compete for orders. If an agent agrees to arrange a dinner for $300 and finds a restaurant that can do it for $270, the network can keep the difference; it can also charge the restaurant a referral fee. These transaction revenues can be used to cover the reasoning costs for users.

However, users should know how agents charge and how they make choices between quotes. Whether they are willing to entrust purchases to agents depends on whether they trust these choices.

Software Can Be Copied, But Networks and Trust Are Hard to Replicate

AI makes it increasingly easy to copy software, while the willingness of people and businesses to transact through a particular network becomes a more important source of advantage. Some of the best companies in the internet era, such as Facebook, eBay, and Uber, become better the more people use them.

Agent business networks can also form this advantage. Competitors can create similar software, but they still need to bring buyers and sellers together and ensure reliable transactions. Returning to the restaurant example, this means there must be enough local restaurants willing to quote and enough groups ready to place orders.

Ensuring reliable transactions is part of building the network itself. Restaurants need to know that agents have the authority to use the group’s money; diners need to know that accepting a quote will lock in a table at the agreed price. The network must be able to verify agent authority, record reservation terms, and arrange payments. Both sides should also know what happens if the group cancels or if the restaurant fails to fulfill the order.

Entrepreneurs can build these services while building the market. As buyers and sellers transact, the network can accumulate records of who fulfills agreements and who does not, forming a dispute resolution process. A network that helps participants find good quotes and reliably complete purchases will bring them back for the next transaction.

Beyond Dining: Travel, Procurement, Computing Power, and Arbitration

Dining is just a toy example. Agents can help build many networks.

In travel, personal agents can coordinate a group’s schedules and budgets, then negotiate a suitable plan with hotels and other suppliers. The more flexible the dates or destinations, the better the quotes they can obtain that fixed searches cannot find.

In corporate procurement, corporate agents can be online all day, negotiating future supply and price commitments with suppliers. They not only compare prices but also delivery schedules and payment terms, helping buyers lock in costs while providing sellers with more predictable demand.

In computing power markets, agents match workloads and available capacity based on hardware requirements, location, and time. Companies can also purchase future computing power to hedge against computing costs.

In arbitration, parties agree to use a shared arbitration platform to resolve disputes, which is faster than existing legal systems. The more buyers and suppliers that adopt the same set of rules, the more attractive the platform becomes for the next participant.

In physical device networks, agents can buy services from drone networks or data from connected devices. The more providers there are, the larger the network coverage or the more types of services needed by other participants.

In collaborative health research, people aggregate health data to participate in crowdsourced studies. The more contributors there are, the larger the scale of research that can be supported, answering questions that small groups cannot.

Human-centered networks also have opportunities. As intelligence becomes cheaper, people still need to establish identities, endorse others' qualifications, and build trust relationships. More work is happening outside traditional institutions, making credential networks and membership clubs potentially more valuable.