What Is an AI Agent?

**Guido Appenzeller** (0:00)
I almost feel like for all the use cases we've described, there's one element that all agents have, which is reasoning and decision. I actually feel like it's a multi-step LM chain with a decision tree.

**Matt Bornstein** (0:13)
A dynamic decision tree.

**Guido Appenzeller** (0:14)
A dynamic decision tree.

**Matt Bornstein** (0:15)
Yeah, I think that's fair.

**Yoko Li** (0:16)
I think we've all just been nerd sniped. I just think, you know, we're computer scientists. So I think we're not well equipped when it's a bit, isn't just zero or one, it's maybe something in between, and we just talk about it a lot. We like try to like coerce it to one value or the other.

**Guido Appenzeller** (0:31)
Yeah.

**Derek Harris** (0:32)
Welcome back to the a16z AI podcast. I'm Derek Harris. It's been a while, but we're making up for the gap with a fun and insightful discussion about what AI agents actually are, including how we should define them, how we should think about the jobs they do, and how the company's building them should think about pricing them. Beyond that, discussion covers things like the distinction between agent workflows, API calls and functions, and the yet unanswered question of how login walls and data silos might affect the ultimate impact of agents. The episode features a16z partners Guido Appenzeller, Matt Bornstein, and Yoko Li, and you'll hear me throughout to smooth the transitions between topics. We had a lot of fun recording this, and you'll hear it all, starting with the question of whether there really is a uniform definition of AI agents after these disclosures. As a reminder, please note that the content here is for informational purposes only, should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any a16z fund. For more details, please see a16z.com/disclosures.

**Matt Bornstein** (1:42)
So I think there's some things which are probably kind of easy to say, which is that, A, there's a good amount of disagreement what is an agent. We've heard a lot of different definitions of it, both on the technical side, as well as say on the marketing and the sales side in some cases, because there's some sales models associated with it. So let's start with the technical side. I think there's sort of continuum here. You know, the simplest thing that I've heard being called an agent, especially just a clever prompt on top of some kind of knowledge base or some kind of context, that has this sort of a chat type interface. So from a user's perspective, this looks like an human agent would look like, right? So for example, I ask it, hey, I have a technical problem with my product XYZ. It looks at the knowledge base and comes back with a canned response.

**Guido Appenzeller** (2:29)
Well, there doesn't have to be a knowledge base, right?

**Matt Bornstein** (2:32)
It doesn't even have to be a knowledge base. I see. Got it. Okay. So maybe it's just a trained model. It's all in the model way. It's the knowledge. So it's even simpler. So an agent could just be an LLM, but the chat interface or something like that, by some definition, right? I think on the other end of the spectrum, there are some people who basically say, for something to be a real agent, it has to be something fairly close to AGI, right? It needs to persist over long periods of time. It needs to be able to learn. It needs to have a knowledge base. It needs to work independently on problems. If you take them the most extensive definition, is it fair to say that doesn't work yet?

**Guido Appenzeller** (3:04)
I think so. It doesn't work yet.

**Matt Bornstein** (3:08)
Will it ever work?

**Guido Appenzeller** (3:10)
That's a philosophical question. All right.

**Matt Bornstein** (3:12)
Fair. Very fair. Very fair. So if we take that continuum in between, is there at least a way to chop that up into a couple of categories of maybe degrees of agentic behavior?

**Guido Appenzeller** (3:26)
And different types of agent. There's some artsy agent that help artists to come up with new, busier curves. There's coding agent, which we like to talk about as the agent of the day. Yeah, which we use. There's agent that's just a rapper on top of LLMs.

**Matt Bornstein** (3:42)
That's right.

**Yoko Li** (3:44)
I may be the contrarian in this group. All right. Look, I kind of think agent is just a word for AI applications, right? Anything that uses AI kind of can be an agent. Now, before we started this talk, I actually went online just to refresh myself about some of the more interesting AI agent perspectives out there. I found a really cool talk from Karpathy that he gave a couple of years ago about agents, which I can describe a little bit. But the really funny part was on the YouTube recommended videos to watch next, it's like AI agents are going to revolutionize your lifestyle and the rise of super-intelligent AI, it's just like marketing. I actually do think that's what's going on a lot of ways. The cleanest definition I've seen of an agent is just something that does complex planning and something that interacts with outside systems. The problem with that definition is all LLMs now do both of those things. They have built-in planning in many cases, and they at least consume information, at least from the Internet, maybe from some servers that expose information through MCP or some other protocol. The line really is very blurry.