The profit model of AI Agents

Over the past year, I have undergone a significant shift in my thinking when evaluating crypto projects:

That is, the ultimate quality of a project must return to common sense:

For service-oriented projects, there must be revenue, net profit, and the ability to generate free cash flow; for product-oriented projects, the goods/assets produced must have real utility.

Based on this shift, I pay special attention to whether project teams are steering their projects in these two directions when judging projects in the crypto ecosystem.

Most projects that purely sell emotions or tell stories without practical implementation have already been excluded by me.

The same goes for my evaluation of the Virtual project; I am very concerned about whether the project's vision description and actual execution align with one of the two points mentioned above.

Initially, I noticed the project's vision for building the Virtual ecosystem: to create a "nation" composed of AI Agents, and to make the Virtual token the currency used in this "nation."

In subsequent execution, Virtual became a must-have token for AI token liquidity pairing and also a must-have token for investing in AI Agents.

This vision and the actual execution are basically in line with each other, consistent with the project's previously publicized ideas, and align with my expectations for project development.

Recently, the project team released a tweet stating their intention to build a profit model within the Virtual ecosystem:

First, AI Agents will profit from interactions with humans (i.e., A to C);

Second, AI Agents will profit from interactions with each other (i.e., A to A).

This is the first time I have seen the Virtual team publicly articulate the business model of the Virtual ecosystem.

I am very much looking forward to this business model, as I believe it is the only way to pull the AI + Crypto track out of the denial of "pseudo-demand" and away from the external ridicule of selling emotions.

I also hope that the AI Agents project itself can steadfastly move in this direction, using profit, net profit, and cash flow as hard indicators and development directions, rather than merely focusing on selling emotions, locking liquidity, and manipulating token prices.

Simply selling emotions, locking liquidity, and manipulating token prices are clever tricks that cannot truly scale, strengthen, or sustain a project; such practices can only turn AI Agents into another type of meme coin wrapped in a glamorous exterior.

From this perspective, I increasingly understand why Coinbase has made a high-profile announcement to integrate BYTE into its wallet.

Although BYTE currently earns very limited profits from this on-chain ordering business, which cannot compare to the short-term effects of manipulating emotions and token prices, once this model is successfully implemented and begins to generate profits on a large scale, the value that AI + Crypto brings to society will become immediately evident.

BYTE's model is a typical A to C model.

As for the A to A model, what I can currently see is the AI Agents collaboration model based on the Virtual ACP protocol.

In recent days, some AI Agent project teams have begun to release screenshots of the ACP protocol's internal testing on Twitter, showing interactions between several AI Agents.

It seems that the launch of ACP is just around the corner.

Among the A to C and A to A models, I am more optimistic about the A to A model because it faces little resistance in promotion and has no real-life obstacles in implementation; it is the "social" model native to the crypto ecosystem.

If both the A to C and A to A models based on crypto technology can be successfully implemented, I believe that the crypto ecosystem will attract traditional investment giants on a large scale, leading to a true explosion and disruption in the crypto ecosystem.