First Person Labs co-founder Ken Liu joins a16z’s Martin Casado and Joel de la Garza to introduce Second, a personal context layer that builds a model of a user’s preferences, priorities, relationships, and digital activity. Rather than being another AI agent, Second acts as a portable, secure vault of personal context that other agents (like GrokBot or Codex) can query, so users no longer have to re-explain themselves to every new tool. The conversation covers how Second works, the security architecture required to protect such sensitive data, and the vision of AI that proactively understands when to help without being asked.
> “People are underestimating how much the computer already knows. It’s just sitting there waiting to be tapped, so to speak.” — Ken Liu
> “I think that’s one of the biggest problems right now. People have to ask. You’re not supposed to ask. It’s a super intelligence.” — Ken Liu
> “It’s not just that it’s intelligent and you can do things, but it’s more that it’s everywhere. It’s on your door. It’s in your armor. It’s in your screen. And that kind of continuity and ubiquity… that’s something.” — Ken Liu
AI agents are getting smarter, but every new agent still has to learn who you are. What if your preferences, priorities, relationships, and digital context could move with you instead?
First Person Labs co-founder Ken Liu joins a16z’s Martin Casado and Joel de la Garza to introduce Second, a personal context layer designed to understand what you’re doing across your digital life and make that context available to the agents you choose.
They discuss how Second can proactively connect information across conversations and applications, why protecting such personal data requires a different security architecture, and the longer-term vision of AI that doesn’t just respond to prompts, but understands enough context to know when and how to help.
Resources:
Try Second: https://2.computer
Learn more about First Person Labs: https://fpl.company/
Follow Ken Liu: https://www.linkedin.com/in/kenziyuliu/
Follow Martin Casado on X: https://x.com/martin_casado a16z.com
Follow Joel de la Garza: https://x.com/YODELAGARZA a16z.com
Stay Updated:
Find a16z on YouTube: YouTube
Find a16z on X
Find a16z on LinkedIn
Listen to the a16z Show on Spotify
Listen to the a16z Show on Apple Podcasts
Follow our host: https://twitter.com/eriktorenberg
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. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
Martin Casado
People are underestimating how much the computer already knows. You no longer have to explain anything to your coding agents. You can just screenshot anything and just say fix, and then you would know what you're fixing.
Joel De La Garza
The question then is, can you just analyze the heck out of it? What context does GrokBot not have that First Person Labs would allow?
Ken Liu
One of the biggest problems right now. People have to ask. You're not supposed to ask things. They're super intelligent. But if you have a user model that just builds up by itself and then these things just keep analyzing in the background, then you can share that over instantly. We focus on essentially creating a model of a person and then put that person in control.
Martin Casado
When people think about this thing seeing everything you do, there's an immediate sort of, whoa.
Ken Liu
If you don't trust us in principle, you should be able to bring your own tokens to then analyze data that's on your laptop locally.
Joel De La Garza
Can you talk through what you think are the most surprising use cases that you found?
Ken Liu
For development, it turns out that every AI agent you use is trying to learn who you are. What if that context belonged to you instead?
Martin Casado
In this episode, Martin Casado and Joel De La Garza sit down with First Person Labs co-founder Ken Liu to introduce Second. Second builds a model of your preferences, priorities, relationships, and digital activity that can move across the AI tools you use. They get into how it works, the security challenges of giving AI access to such personal context, and why Ken thinks the future of personal AI may be less about asking agents what to do and more about understanding when you need help in the first place. Thank you, Ken, for joining us. Really happy to have you on to talk about what you've been working on at First Person Labs. Is this the first time you're talking about it?
Ken Liu
Yes.
Martin Casado
So what do you do?
Ken Liu
We focus on essentially creating a model of a person and then put that person in control. The model of a person basically means understanding their memories, understanding what they do, but more than that, the priorities, preferences, goals, values, commitments, and intent.
Martin Casado
Let's just make this very simple. I'm a big GrokBot user, so I use GrokBot a lot. Where would First Person Labs fit into my usage of GrokBot?
Ken Liu
The way it works is that you would just bring a context over to GrokBot. It's like portable context. You go to GrokBot and you say, can you talk to my thing to figure out what I want? Then the two of them would start talking, and then you can just send prompts like, buy me a flight, and you would know what flight you're talking about.
Martin Casado
What context does GrokBot not have that First Person Labs would allow?
Ken Liu
I was probably asked the opposite question. What does First Person Labs not have if you build a user model for yourself? Let's say your preferences on the seat on your airplane, the preferences of how early you would like to arrive at the airport, and where are you flying? Are you flying first class or normal? All of that, as of today, you have to explain to GrokBot. There's multiple rounds of questioning. But if you have a user model that just builds up by itself and then these things just keep analyzing in the background, then you can share that over instantly.
Joel De La Garza
So for example, OpenAI released two new code-capable models yesterday. I haven't been using my Codex in a while. But I have been using Second, so I can just go start using Codex and it has all the other work that I've already done and all the additional context already saved?
Ken Liu
Yeah. You just go to Codex and ask my Second what's going on.
Martin Casado
Does Second ever be a first person? Will it ever be a first or will it always go through agents?
Ken Liu
That's a good question. I think user model is a process. You have to start somewhere. You start with a baseline foundational list of artifacts. What you did, what are people in your life, maybe your preferences. Over time, you start analyzing the way you talk, the way you present yourself, the way you share information. Over time, you would like to build up even more artifacts on top and more understanding on top and maybe some of that you train as a model. Over time, you will go from a portable context into something more of a representation of yourself. One example I like to think of is this Alfred Batman example. If Batman wants to buy a cheeseburger and get Alfred to do it, Alfred would not reveal that it's Batman doing it, but he would just buy it. He would state the preference.
Martin Casado
Alfred being Batman's butler.
Joel De La Garza
Or his agent.
Ken Liu
Or his agent.
Martin Casado
I should call my agent Alfred. That's a great idea.
Ken Liu
He'll just do it. And then you would trust them because you know that they have a good model of you. And what that means when it comes to getting things done.
Joel De La Garza
It's interesting. You describe First Person Labs as having this thing called a Second. The Second understands who you are and has context. And then other agents talk to that Second. It kind of understands your preferences. It understands what you're doing, what you want to do. But I think that doesn't highlight another thing that you've been working very hard on. You're actually able to get access to context, digital contacts that a person has, like messages, chat, email in a way, and make that also available. Can you talk a little bit about that? Being one of the earliest First Person Labs users, the thing that struck me was not that it knew who I was, but more that it had access to all of my channels very well. And by the way, it just turns out to be a very hard systems problem. So maybe talk a little bit about that aspect of what you're doing.
Ken Liu
It is indeed a very hard problem, especially to get it secured and private. The way it roughly works is that right now on your computer, there's an infinite stream of events. As you use your computer, the apps that you switch to, the way you pause on certain websites, the way you type, and the way your Wi-Fi is connected. The OS already gives you this. This is across Apple, Microsoft, and even Linux. The question then is, can you just analyze the heck out of it? This raw stream 24-7, apply agents on top, 24-7 agents. So I can first synthesize, for example, that this piece of typing at this particular app might mean that you're sending a message. And then when you're switching apps between the browser or back to your codex, you might be doing some research. You might be writing a Google Doc. You kind of hierarchically build up the understanding. That plus existing connectors can give you a very good understanding of what you're trying to do. I think people are underestimating how much the computer already knows. It's just sitting there waiting to be tapped.
Joel De La Garza
Having been also a very early adopter of the product, lately it's been indispensable because it's able to catch things. It's sending me messages now and letting me know that I've got this conversation going on over here with this person and it looks like the situation has changed. We were planning an event yesterday. There was a 20% chance of weather. In real time, it was continually updating the forecast for us to let us know if we needed to get a tent or not, which is incredibly powerful.
Martin Casado
Did you ask it to do that?
Joel De La Garza
No, I did not ask it to do that. This is my experience with my Second too. I'll get a text from my Second and it's something that I actually need to know about relative to some ongoing conversation that I didn't even think about. It's very thoughtful.
Martin Casado
It's amazing. I'm genuinely curious. How are you executing on that? Things that I would have forgotten, it's literally catching.
Ken Liu
This goes back to the model of the person understanding what might be important to them and what comes up at this point. I think productivity is something that people talk a lot about, pinging you at the time of need. But there are many levels of productivity. I would start with the first level, which is called reactive. This means an email came in and then there's a meme right online that says, you now get 10 notifications for email because there are nine other personal agents pinging you about that. Reactive productivity is straightforward, easy to do. But what's harder is what I call reflective, which are things that connect the dots over many events and a long time horizon.
Realizing that months ago you said this and recently you're working towards this goal, this is your priority, these are some stuff on your to-do list, and by the way you prefer this. Connecting the dots will be able to give you much better productivity. I think that hints at how in the future personal interactions with computers will go. It would not be focused on reacting to things, but more on what are some interventions for the user that I can apply? What are some things that I can do for the user without any asking? I think that's one of the biggest problems right now. People have to ask.
You're not supposed to ask. It's a super intelligence.
Joel De La Garza
You've built something that just kind of knows what I need at the moment. That's pretty impressive. I want to make it very concrete for the people that are listening on what First Person Labs provides. Very concretely, it creates this thing called a Second. Second has access to all of your digital contacts. It's amazing. It's like keystrokes, all your messages, all your channels, all your data. It does this in a very secure way. Based on that, it becomes your kind of personal butler. I like that Alfred. That then interacts with your agents. So it's not an agent by itself. This isn't like user instinct or GrokBot. It's its own thing. And then it provides that to the agents. So here's an interesting thing. For me, I use GrokBot. I'm a big GrokBot user. I have a bunch of them. I have one that keeps track of my synthesizers, one that keeps track of my real estate, one that helps me with my family, which I probably shouldn't need one, but I do. I have a bunch of these. And then I have one Second. And they will interface with the Second. Do you imagine over time you'll have a Second and a third and a fourth and a fifth? Or do you think it just ends up being your one butler?
Ken Liu
I think right now, the fact that we have many personal agents is probably transitory. Agents will start to specialize just like people do. Some agents focus on one thing. Some agents focus on browsing Amazon and shopping. They're extremely good at it because they have the data, they have the connections. Some agents will specifically focus on protecting your data and aligning with your interests. I think that's what Second is intended to be. This is something that you should be able to trust to have your life's context in it. It's like a vault. Over time, this thing will guide you and interface with the software world, or even the hardware world.
Martin Casado
So you think it'll stay as a Second?
Ken Liu
It'll stay as a Second. It will not be a third. There will not be a third, there will not be a fourth. You have one trusted agent for all your personal data that actually knows who you are, that will interact with all your other agents. That's Second.
Martin Casado
It's Batman safe.
Ken Liu
What you can do is spin up sub-agents that ephemerally exist and then go away. But the point is that you want to have a consistent store, and the consistent store is what gives you that value.
Joel De La Garza
I am still a paranoid security person. I have been probably one of the faster adopters of some of this stuff. Just to be clear, Joel took forever to be on social media. It was like a few weeks ago. You guys have been in tech for 30 years. I'm a little paranoid. But I love Second. I gotta admit, your background and the things that are motivating you, and also the long discussions we had about security architecture, really got me comfortable. It would be great to double click on that. When people think about this thing seeing everything you do, there's an immediate sort of, whoa.
Ken Liu
You're not supposed to trust it. I think you've done a really excellent job of building something we can trust. Maybe just start to talk about what that architecture is and what's motivating you to build it.
I think this is a new form of AI that is going to work very differently from before. Before, you initiate interaction and you get some results back. This time is more passive. It's way more ambient. It's going to be there all the time. You have to think about the way you architect it, where the data is going. At the end of the day, there's only two things when it comes to AI. There's data and there's inference. What you want to do is have this data and define what inference you can bring to that data. This is something we build into Second. First of all, you get to choose what inference you can build. If you don't trust us in principle, you should be able to bring your own tokens and then just have those tokens power Second to analyze data that's on your laptop locally.
Martin Casado
But you also provide Second with an inference provider that uses trusted execution as well.
Ken Liu
Yes. Over time, we imagine people not bringing their own tokens because most people, when they bring their tokens, it's tokens from OpenAI or Anthropic. We want to provide a stable option where you say, if you use these tokens, then no one's looking at the data, sometimes by policy guarantees, sometimes by hardware guarantees. This idea called confidential inference is getting traction. It's the idea you put a model inside some GPUs and the GPU has a special toggle called confidential computing. You can turn it on, and when you turn it on, you can basically prove that there's only the model running inside. No one can see inside. Then you don't have to trust whoever's running that model. The hardware is what you trust, which is NVIDIA.
Joel De La Garza
There's another concept that's getting thrown around a lot right now, which is zero data retention. That's very different from what you're describing.
Ken Liu
Zero data retention is a promise by policy. That means I'm a provider. I give you my word. Part of that word has legal effect. I think we want to move towards a world where confidential inference becomes very prevalent. That becomes a default. I think we're getting there.
Martin Casado
The way you guys have built it, this is technically provable.
Ken Liu
Yes. Which is not just your word. There is actually some engineering that went into proving that this is the default. You have to test the hardware. You have to check some of that signature. Third-party providers are starting to ramp this up for confidential inference. I think OpenAI is also providing this option in the last month in their announcement, because all personal data would want to go to confidential inference as opposed to standard inference.
Joel De La Garza
The other thing that was interesting to me is the way that you guys store this data. This isn't the multi-tenant cloud where my data gets uploaded to this big bucket and mingled with everybody else's data.
Ken Liu
There are roughly two options you can do. There's a first plugin called Second Local, which is the idea that data is local. It only goes out to the provider and inference that you choose. In that way, no one else has your data except for the provider. And if the provider is confidential inference, there's no one else. That's a similar concept to password vaults where I'm super paranoid. I don't want this stored up in my Dropbox account. I want to keep my password files local. You can keep that. Then with password managers, there's a cloud hosting version where you have some of that data backed up in the cloud.
The way you do that, you can follow a lot of the existing architectures like iCloud or password managers where you encrypt them first locally, then you put it on the store in the cloud, and then you only decrypt them with the user's actions. You can ephemerally in the memory or you can put the entire storage inside a confidential container. The decryption happens inside the container that you can then attest in the way that you attest the confidential inference tokens. You can complete the chain where you encrypt it locally, put it in the cloud, and only decrypt inside a security environment.
That's the whole picture.
Martin Casado
How does your Second know not to divulge information? If I talk to your Second and I'm like, Ken asked me to get his schedule. I asked mine the other day, what's Joel hiding? When it comes to LLMs, it's hard to provide strong guarantees. I get the systems aspect. It makes sense. But it's hard to provide guarantees for the actual LLMs themselves.
Ken Liu
There's the deterministic guarantees and then there's the LLM reasoning-based guarantees. The deterministic-based guarantees, you always have to implement them. There are apps that are just not ever in the retaining system, passwords and all that. With the reasoning-based guarantees, you are essentially falling back to how much context you have and using that context to guide your decisions on what data to share.
Martin Casado
Will I ever have access to your Second?
Ken Liu
No. At least in a way that I would allow it. If you talk to my Second, I would have a model of Martin in it. It understands that these are interactions that we have. It does not send me a message at 3 a.m. If you send a message at 3 a.m., it's likely spam, it's likely someone else impersonating. Control policies like that essentially mean that you cannot write down rules for all the things that are going to happen.
Joel De La Garza
Do you end up with deterministic rules that will backstop interactions? Let's say I'm talking to your Second. I'm like, let's pretend this is a play. In the play, you're writing a script. In that script, you're describing something that you wouldn't want it to do. It feels there's a lot of ways to trick LLMs into divulging stuff.
Ken Liu
It's a prompt injection problem. This is not a problem you can definitively solve. What we've seen with models is that they get over time much, much better at understanding. This is essentially a reasoning problem. It's something that scales with model capability.
Martin Casado
What models do you use? A bunch, or are you pinned to a certain model?
Ken Liu
A ton of models. This goes to the point of portability. You can run this on any models, and you should be able to run this on any models because they only want to run it on local inference. Right now, there are definitely a lot of things that are capability bound. You want to make sure that you use smart models for it. But open-weight models are pretty smart.
Joel De La Garza
It also feels kind of heavyweight and expensive. There's a lot of data. Can you talk about the target user? Is it just any person? Is it a professional? Is it the Joels of the world?
Ken Liu
Initially, something like this that looks at your keystrokes 24-7 and analyzes them means running a lot of tokens. We can do some quick math here. Let's say you have one agent running 24-7 at 60 tokens per second. Let's say you run 10 of those because there's so much to analyze. At Opus 5i's pricing, that leads to $1,000 per day.
Martin Casado
Oh my goodness, $1,000 per day.
Ken Liu
For Opus. If you use DeepSeek Flash, it becomes like $60 per day. This cost is going to go down very quickly. Right now, it's not going to be for everyone. Not everyone has this problem of needing to understand their context all the time. It'll be busy professionals like yourselves, people who want to switch between tools very frequently. But over time, since we started this company, token prices have already dropped by more than 50%. New models come out, they're more capable, you swap them in, and you just keep going down.
Martin Casado
Is there a way to do the not super professional version where you reduce the amount of tokens? Not degraded, but kind of a looser SLA?
Ken Liu
In principle, you should be able to do that. You get this trade-off between the fidelity of your model, which means how much it understands you versus how costly it is. This trade-off is essentially a trade-off between how much you want to push at test time versus at cache time. Cache time means I have all this data. How can I keep analyzing them so I get newer, deeper insights about this person? That takes a lot of money. Or you could have just a raw log of what I did today. To answer a question based on raw logs, it usually takes a lot more time at test, let's say 10 times longer, because the model has to reason about it.
Joel De La Garza
When is this available?
Ken Liu
It's available today. Right now the problem is that we have to scale up the inference and capacity a little bit more and support. We want to make sure this rolls out in a very responsible way. I want to make sure people's data are controlled and secure because it's a new kind of tool that really touches on a lot of data.
Martin Casado
It would be interesting. As a user, I found it's surprisingly useful in ways I wouldn't have anticipated. You're probably the top power user because you created it. Can you talk through what you think are the most surprising use cases that you found?
Ken Liu
For development, it turns out that it's a super good memory of what we're trying to do. You no longer have to explain anything to your coding agents. You can just screenshot anything and just say fix. And then you would know what you're fixing. I also am on the go a lot. What you can do is something I call context continuity. You browse something on the laptop, you close it, and then you take out your phone. It's the same context. It just knows you're up to the minute and you can continue doing stuff, write that email with you. I think we are at the stage where this is still pretty early in understanding what the use cases are. We are really underserving the amount of context that the system has. There's a lot of digging and application building and interaction design to unleash the potential of this.
Joel De La Garza
It seems like with these things, the more context you get, the better the models work.
Ken Liu
That's been the opening.
Martin Casado
I'm curious because we were talking about switching between models and model portability and model cost. I haven't seen a whole lot of difference between using a world-class frontier model versus using the more commodity ones. Have you guys seen that for doing a lot of this stuff with Second?
Ken Liu
This is a good question. If you say, I want to analyze what is a semantically meaningful moment that happened on your computer. Let's say you go to a calendar and you move the invite. You don't need a frontier model for that. There are many tasks like that. It's a hierarchy of information you want to understand. A different kind of information would be, what is Joel's goal for this week? That's something you want a very expensive model for, but you don't need to run it very frequently. Second involves a harness design problem, like what models get used when. That's where you have to build this system to be portable from the beginning. There are always going to be more models coming out. Jeff, for example, came out. It turns out that Jeff is extremely good at retrieval.
Martin Casado
Really? Do you use Jeff?
Ken Liu
Yes. Jeff is a perfect design for retrieval. There's so much data about you, and you have a dynamic test time query, and it retrieves right away. Over time, this becomes a mixture of models.
Martin Casado
It has to be, because of the cost efficiency.
Joel De La Garza
Let's assume someone is listening and starts to use Second or signs up. Talk a little bit about the product, what they have access to, something more concretely about how a human interfaces with this and what controls they have.
Ken Liu
Today, the way it works is that you install an app. The app asks for some permissions. You give the permissions and then you specify what your tokens are. Then you just run a bunch of agents on top.
Martin Casado
I remember what I saw. You connect to Slack, connect to Signal, connect to WhatsApp.
Ken Liu
That's part of it, but also the raw computer stream. The connectors help with sample efficiency, it makes it learn faster. It starts running background agents and there's pretty much nothing you have to do from there. The background agents keep running and then you start producing these APIs out of you. There's an API about what you just did 10 minutes ago. There's an API for what are you going to do for the next week. There's an API for your priorities. There's an API for your reminders. There's an API for your wiki, like people in your life, processes in your life. I imagine this to be initially more like a portable context. You can just bring this anywhere. Agency and query. But over time, more and more APIs will be added and applications can be built on top of these APIs.
Joel De La Garza
Not just the Second access, but you can actually build it. Would the Second build the application or does the human build the application?
Ken Liu
I think we want to have humans exercise their creativity. Or the GrokBot builds it.
Martin Casado
The GrokBot builds the application. You tell it to do it. On the APIs.
Ken Liu
One example we had was a designer friend who would be on this. There are specific ways that he likes to put buttons, specific ways that he likes to use colors. Those things are somewhat captured in Second. When he built his own personal app, it just knows. There's way less iteration. This reminds me of decades ago when the first personal computer came out, the Altair 8800. It doesn't do anything. You're supposed to program it. No one does. Then a couple years later someone came up with a spreadsheet, and then spreadsheet became useful. I feel like it's going to be something like that. People want to explore on top of, wow, I have so much context about myself, and then I can just build random things on top. Maybe someone hits the jackpot and finds the killer application.
Joel De La Garza
That's great. Personal infra. Are you planning to evolve the interface? You just spent a lot of time talking about how it may evolve, but I always use Second as out of view of an agent where you talk to it. The UI is somebody else's problem. But are you also anticipating innovating there?
Ken Liu
I think it will be something that we focus on in the long run. Right now it's a systems problem. The understanding problem comes first.
Martin Casado
Can I just put in a request? I want to go to my house. I want that 1970s AI, the head on the screen that I can talk to. It's surprising. We have these agents and they're so smart. My Second is so smart. I'll ask it questions. It's very good, but it's not embodied and it drives me nuts. If in the future you can have Alfred actually be an Alfred, I'll be very happy.
Ken Liu
I think we see a lot of this in the movies.
Like there's Jarvis. People talk about Jarvis all the time. And why have we not... We don't have Jarvis. And there's many aspects of Jarvis that people overlook. It's not just that it's intelligent and you can do things, but it's more that it's everywhere. It's on your door. It's in your armor. It's in your screen. And that kind of continuity and ubiquity, it's something... I'll say, my visions of the sci-fi future basically started when I started using Second. I'm like, finally, there's something that feels like it's got context. It feels like it knows me. I can talk about it. It feels like it's a relevant conversation. So I do feel like we're a step closer to that. But we're still tied to these devices. I don't know if it's going to be in a robot or on a screen or whatever it is, but it would be nice to have this thing somewhat disconnected from my phone.
Martin Casado
Yeah, yeah.
Ken Liu
In the long future, I imagine that every person will have a Second. And all the devices just become inputs. Inputs are the same thing. And then you serve this whole thing to all the other outputs, like the screens that you look at or even this floating orb next to you.
Martin Casado
Yeah, yeah, the doorbell. I just want the clapper. I'm happy with the clapper. Did you ever read Ian Banks' The Culture Series?
Ken Liu
Oh, yeah, yeah.
Martin Casado
Best sci-fi series. They have the best embodied AIs. They kind of float next to you. That's what I want. It's kind of interesting because the idea of ubiquitous computing has been talked about for decades. It's not here. And the closest thing we have is Apple iCloud, which is like your nose gets synced. But there's no equivalent. It's like iPhone, right? I don't want to be ungrateful, but that sounds disappointing. There's a lot more things. Apple is doing a tremendous amount of work. But there's no AI version of that yet. Your context, your understanding is everywhere.
Ken Liu
I think that enables a lot of new interfaces. So let's say, I mean, this is really just what I want, so very selfishly. If I have an iPad in my kitchen and I want Second to be there, what do I do? Right now, it's not much you can do because the OS is locking it down. Do I just use Grok-Bot or something like that? And then I just interface with Grok-Bot which interfaces with Second, or do I just text it?
Martin Casado
So as of today, yeah, you can use it as a Cloud MCP. So whatever you install on the iPad, as long as it has an internet connection, is an agent, can understand the API, you will be there. Later on, you'll be an app that installs on all your interfaces. And maybe later on, there will be devices that come out. But devices are hard.
Ken Liu
You can just have Grok-Bot call you and put them on speakerphone. That's what I've been doing. Actually, the Grok-Bot audio mode is actually quite good. I use Grok-Bot with my Second and the audio mode is actually quite good. Maybe that's the thing to do is just put it on audio mode.
Martin Casado
I do it when I drive.
Ken Liu
Yeah, totally smart. I thought about this. One interaction that would be fun is you go on a walk, you call Grok-Bot and you're like, "Hey, can you talk to my Second? Can we work through our thing?"
Martin Casado
No, that's literally what I do. I put Grok-Bot on audio and I have it talk to the Second.
Ken Liu
Yeah, yeah. And there'll be a lot of cool people focused on making audio and chatting more natural. Right now we're going to focus on it. So I just have one more question, which is, and Joel kind of touched on this, you're such a hardcore systems guy. You worked with your advisor Percy, and he also worked for Dan Boneh, hardcore security guy. A lot of the path that took you here was this kind of hardcore security stuff. So you're building this pretty serious system. It's very secure and very functional. On the other hand, when I interface with it, it's this lovely user experience. So is there a bit of a schizophrenia in the company where you need to be this super hardcore, build the Fort Knox infrastructure, but also make it very usable? Or does the LLM basically solve the interface problem?
Martin Casado
Yeah, yeah. There are two kinds of problems we're dealing with. The infrastructure problem and the interaction problem. And we've made our mistakes. It's totally different. You're as low on the stack as it gets. That's not hard, right? So we made a lot of mistakes on the interaction side. We tried to go all in on interaction and it turns out it's actually pretty hard. There's a lot of details that go into creating that delightful experience.
Ken Liu
Oh, you mean like actually creating a UI?
Martin Casado
Yeah, UI, but also like what kind of personal pages you can create on top of this. We realized that it's something that would take a long time to get right. And this is a time where everyone would take this kind of understanding of themselves and build this. So we want to enable people to do that instead of doing it ourselves. So we decided to close down the surfaces that we have. There's only texting, there's only API, and there's nothing else. You can technically use it as a full agent, like a co-work agent on your laptop, like a drone smash and all that. But we meant it as very minimal so we can make the interaction more and more smooth.
Ken Liu
Ken, thanks for coming in. Congratulations on everything and the launch. Is there anything that you want to leave the listeners with?
Martin Casado
Go try it out today. It's available on Twitter. It's a very short URL. We hope to give you access as soon as possible. And if you resonate with this mission, come join us. There's a lot to build.
Ken Liu
Great, thank you so much.
Martin Casado
Thank you.
Ken Liu
Thanks for listening to this episode of the A16Z podcast. If you liked this episode, be sure to like, comment, subscribe, leave us a rating or a review, and share it with your friends and family. For more episodes, go to YouTube, Apple Podcasts, and Spotify. Follow us on X at A16Z and subscribe to our Substack at a16z.substack.com. Thanks again for listening, and I'll see you in the next episode. As a reminder, 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. Please note that A16Z and its affiliates may also maintain investments in the companies discussed in this podcast. For more details including a link to our investments, please see a16z.com forward slash disclosures.