Decoding AI Agents: Clarity Amidst the Hype

Cansın Cengiz

Cansın Cengiz

— 3 min read

What *Really* is an AI Agent? Unpacking the Buzz with a16z VCs

The term "AI agent" is everywhere, but what does it truly mean? It's a buzzword stretched so thin that even top venture capitalists at Andreessen Horowitz (a16z), a firm heavily investing in AI, admit there's no agreed-upon definition. Let's dive into what they think and what it means for the future of AI.

The Quest for Definition: Insights from a16z

During a recent podcast episode, “What Is an AI Agent?”, three a16z infrastructure investment partners—Guido Appenzeller, Matt Bornstein, and Yoko Li—attempted to nail down a definition. This is significant coming from a16z, which has backed AI giants like OpenAI and Anysphere (maker of Cursor) and is reportedly eyeing a $20 billion megafund for further AI investment. Their perspective is crucial, as they've previously stated on their corporate blog, “We believe every white-collar role will have an AI copilot. Some of these roles will be fully automated with AI agents.”

Appenzeller noted a "continuum" of AI startups now claim their products are agents. “The simplest thing that I’ve heard being called an agent is basically just a clever prompt on top of some kind of knowledge base,” he said, describing systems that might offer canned responses for IT help desks.

Human Replacement: Hype vs. Current Reality

The conversation often escalates to AI agents as human worker replacements. Companies like Artisan (despite its “stop hiring humans” ad campaign) and others looking to replace human workers fuel this narrative. However, Appenzeller argues that true human replacement would require AI “something close to AGI,” capable of long-term persistence and independent problem-solving. The current reality? “It doesn’t work yet,” according to both Appenzeller and Li.

Jaspar Carmichael-Jack, CEO of Artisan, confirmed to TechCrunch that developing reliable AI agent tech has been a surprisingly hard journey. Key technical challenges include:

No company wants an unreliable employee, human or artificial, that forgets or fabricates information.

A Working Definition for Today’s Agents

The a16z trio did propose a practical definition for current AI agents. As Li described, an AI agent is a reasoning, multi-step LLM with a dynamic decision tree. In other words, she said, an agent isn’t a bot that just does a task when asked, but it must be able to make decisions about the task and take action autonomously, like grab a list of prospects from a database, decide which ones to email, and write the emails. Or write code and decide where to insert it.

Future Outlook: Augmentation, Not Annihilation

Could agents replace humans in the foreseeable future? The VCs agreed they could be used to handle some tasks humans do now, just like automation has always done. But this may actually lead to companies hiring more human workers, not fewer, as productivity rises.

Bornstein said he can’t envision a time – given the current state of agents – where humans will be unnecessary. From “our perch in Silicon Valley,” the tech industry can “forget” that most people have jobs that require human creativity and “thinking,” he described. To replace humans with a bot, “I’m just not sure that even is kind-of theoretically possible,” he said. Still, such human replacement rhetoric — often done for marketing/business model and/or pricing reasons — is “a big reason for the confusion we’re experiencing now,” Bornstein says.

The Bottom Line: Proceed with Informed Caution

The upshot is, if those seeing all the most cutting-edge uses of AI agents are skeptical about the boldest claims AI agent companies are making today, that’s probably a good sign the rest of us should be, too.

References

Read more

Lex Proxima Studios LTD