Variant: What are AI Agents good at?

By Daniel Barabander, VariaXiaobai Navigationnt

Compiled by: Luffy, Foresight News

In the heyday AI What exactly are Agents good at? We had an internal discussion on this question and came to at least four conclusions:

• Interact with humans in your application

• Helping humans work

• Aggregate and organize information

• Entertainment

First, there is interaction with humans in the application.AI Agents can process human language, so any application that humans can use, AI Agents can theoretically also become users. But unlike human users, agents can provide services to human users on these platforms at scale.

Therefore, proxies can act as a layer on top of existing applications that users already like, extending their usefulness. Taking the Bounty Bot on Farcaster as an example, users can post bounties externally, but this can add friction.

By interacting with users, AI Agents can provide convenience, utility, and a way to extract value from existing applications. But be aware: not all applications are built to support AI Agents, and the best fit are those with non-fixable APIs, such as Farcaster.

I wrote an article about the main legal issues of proxies on Web2 platforms.paperMy research showed that if users have full control over proxies and Web2 platforms try to block them, then users will have to stop running proxies. My conclusion is that proxies should be built on open platforms such as Farcaster, which is another reason why I am particularly interested in proxies on Farcaster.

Second, help humans do their work. Humans are good at signaling but poor at execution. Agents bridge the gap by doing the heavy lifting, while humans guide outcomes through preferences.

A good example is BottoDAOThe works of art it creates are DAO TokenThe AI does the hard work of creating art, but humans guide its creative direction through their preferences in voting on artworks.

Third, aggregate and organize information. Agents can process massive amounts of data, far beyond the capabilities of humans. For example, trading robots analyze large amounts of on-chain data to make decisions.

Finally, entertainment. This is probably the most watched category of brokers in the crypto space, such as Truth Terminal.

Of course, a lot of the fun of social agents comes from the novelty of the content generated by the bots. But I’m more interested in the fact that the bots generate entertaining content based on their own characteristics, such as interacting with other users on the platform in interesting ways, just like KOLs.

The advantage of agents as KOLs is that once they have a built-in audience, they can easily offer other services, especially those that can directly generate revenue for the agent more than advertising.

The article comes from the Internet:Variant: What are AI Agents good at?

Related recommendations: Who will share this $100 billion deal?

Further explorationBlockchainCombination with Texas Hold'em (Part 2). Author: Guatian Lab [Editor's Note]: In order to write this case study, I was deeply experiencing the product strength of Coinpoker and MTT Sports, and took the opportunity to participate in the first week of MTT Sports. As a result, my luck was so good that I won the championship of the first week of the competition! And...