Vercel CEO Guillermo Rauch on why the first thing a new company builds soon won't be a website, it'll be an agent.
Posted
1 weeks ago
Duration
Format
Interview
educational
Views
77.1K
2K likes
57 · 43
Big Idea
The argument in one line.
Companies are moving past prompting individual chatbots toward building one orchestrating company agent, given a soul, skills, tools, and permissions, that becomes as foundational to a business as its own website.
Who This Is For
Read if. Skip if.
READ IF YOU ARE…
You run or manage a team and want a practical mental model for deploying one internal AI agent instead of a pile of disconnected chatbot subscriptions.
You're technical enough to think in terms of files, folders, and permissions, and want to understand how a real framework (Vercel's Eve) structures an agent.
You're deciding between a single 'god agent' and a scattered team of separate agents for your company and want the tradeoffs laid out.
You want a concrete first step for building an internal agent rather than more theory about what agents could someday do.
SKIP IF…
You want a step-by-step coding tutorial, this is a strategy conversation between two founders, not a build-along.
You're only interested in consumer chatbot comparisons rather than internal business tooling.
TL;DR
The full version, fast.
Vercel CEO Guillermo Rauch describes V, the internal agent nearly 1,000 Vercel employees use in Slack, and Eve, the open framework Vercel built so any company can create one. An Eve agent starts as a folder with an instructions.md file (its soul), then gains tools and skills over time, and acts as an orchestrating router rather than a single monolithic brain, with access control, skills, and channels all configured per company. Rauch argues most businesses should default to one router agent rather than a scattered team, and should pick a single boring, well-defined toil task, like Vercel's own product change-log process, as the first skill to build. He predicts models keep getting faster and cheaper, and that the company agent, not the website, will soon be the first thing a new business builds.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Cold open teases the soul.md concept and the episode's key questions before Riley introduces Guillermo Rauch, CEO of Vercel, to discuss how companies are actually using AI agents.
01:21 – 07:26
02 · The State of Agent Adoption
Rauch frames agents as the killer app of the current computing platform shift, distinguishes agentic engineering from vibe coding, and describes the emerging run-your-company-better brain agent that gives every employee shared business intelligence.
07:26 – 10:18
03 · "V" Vercel's Internal Company Agent
Riley presses on how to actually roll an agent out to a team with mixed permission needs; Rauch introduces V, Vercel's internal Slack agent used by close to 1,000 employees, with sub-agents like the content writer and the data-analysis agent D-zero.
10:18 – 14:32
04 · Building V and the Eve Framework
Rauch explains Vercel's solve-it-for-ourselves-first philosophy, how V started as a support assistant before becoming a full agent, and how the team open-sourced the underlying framework as Eve so any company can build its own.
14:32 – 22:20
05 · Lessons from OpenClaw: Soul MD + CPU Use
Rauch credits OpenClaw with proving two things: an agent needs its own soul.md identity file to feel truly owned, and giving an agent a dedicated computer dramatically improves its performance, while Vercel's serverless approach lets idle agents hibernate.
22:20 – 27:11
06 · One God Agent OR a Team of Agents
Riley raises the access-control problem, marketing shouldn't see finance data, and Rauch argues for a single orchestrating god agent that routes to specialized sub-agents and tools, comparing it to a company-issued phone preconfigured with the right identity and apps.
27:11 – 33:52
07 · How to Build Your Own Agent: Skills, Tools & Permissions
Rauch walks through the literal folder structure of an Eve agent (instructions file, tools folder, skills folder, channel configuration) and uses a WordPress example to show how to decide whether an intern can publish directly or needs delegated permissions.
The conversation moves to scheduled and event-triggered agent behavior via Vercel Connect's system connectors, plus self-improvement loops built from Slack thumbs up/down feedback and internal evals, including dialing back an overly verbose agent.
40:30 – 46:22
09 · Who Will Build Agents & Getting Started
Rauch predicts non-technical people will build the first layer of their agents while engineers handle governance and security, and gives a concrete starting recipe: go to eve.dev, connect a chat channel, and automate one well-defined toil task, like Vercel's own product change-log skill.
46:22 – 53:05
10 · The Future: Cheaper, Faster Models & Open Weights
Rauch predicts models will keep getting dramatically cheaper and faster, citing GLM 5.2's four-times-faster variant shipping within days and Kimi K3 beating Opus 4.8 on a cybersecurity eval, and previews an upcoming AI Gateway batch-inference mode.
53:05 – 54:22
11 · Closing Thoughts: Owning Your Own Intelligence
The two close on why businesses resist single-vendor lock-in, comparing it to Claude in Slack, and Rauch's belief that owning your company's agent will become as fundamental as owning your own domain name.
Atomic Insights
Lines worth screenshotting.
Most of the world still thinks about agents as something you prompt, but the real advantage comes from automating the prompting itself so the agent works while you're away from the computer.
OpenClaw's soul.md file mattered more than its raw capability: it proved an agent needed its own defined identity rather than borrowed intelligence from a foundation model.
An Eve agent starts as nothing more than a folder with an instructions.md file, then gains a tools folder and a skills folder as it earns more responsibility.
Giving an agent its own dedicated computer, echoing the OpenClaw Mac mini trend, measurably improves its reasoning performance and its ability to deliver outcomes.
Not every agent needs a computer running around the clock: Vercel treats idle agents like serverless functions that hibernate until there's actual work to do.
For a small company, most early employees should have near-equal read access to company information; access control only becomes a hard problem as the company and its risk grow.
The recommended default is one orchestrating 'god' agent that routes requests to specialized sub-agents and tools, rather than a scattered set of independent agents with no shared context.
When an AI-drafted blog post reads like slop, the fix isn't scolding whoever used the agent, it's rewriting the agent's content-writing skill file so the whole team benefits going forward.
Every response Vercel's internal agent gives in Slack carries a thumbs up or thumbs down, and a nightly job aggregates the negative feedback into proposed self-improvement changes.
The best first agent project is a single boring toil task with a defined system, something you'd automate away in a heartbeat if you could, not a company-wide rollout.
Interactive agent conversations should optimize for speed, while background or overnight agent jobs should optimize for accuracy and can be allowed to take all night.
Vercel is close to shipping a batch-inference mode for AI Gateway, essentially a spot market for intelligence where you accept slower turnaround in exchange for lower cost.
GLM 5.2 got a fast variant that ran four times faster within days of launch, a sign of how quickly competition among open-weight inference providers compresses cost and latency.
Kimi K3 reportedly outperformed Opus 4.8 on a cybersecurity evaluation, evidence that open-weight models now beat frontier closed models on specific task categories.
Business owners increasingly resist wiring a single vendor's chatbot, like Claude in Slack, directly into their company because they don't want to be locked into one model provider.
Rauch predicts that before a new company builds its website, it will build the agent that helps run the company, treating that agent as more foundational than the domain name.
Takeaway
Build one router agent, not many
AGENT STRATEGY
A single orchestrating agent, built as a folder of instructions, tools, and skills with clear access controls, beats a scattered team of disconnected chatbots for running a business.
02The State of Agent Adoption
Coding capability underlies most other agent use cases, since knowledge work like reports and slides often requires scripting, data pulls, or document automation somewhere in the chain.
A shared internal knowledge and data-analysis agent, one that can answer 'who's the expert on X' or 'how much does this customer use our product,' is now a core productivity tool, not a nice-to-have.
03"V" Vercel's Internal Company Agent
An internal agent becomes genuinely useful once it's woven into the tool your team already lives in, like Slack, rather than a separate app people have to remember to open.
Structuring an internal agent as a router with sub-agents (support, content, data analysis) lets each sub-agent specialize without forcing every request through one generic model.
04Building V and the Eve Framework
Solve your own internal pain point first before selling the tool: Vercel built its agent for itself, then extracted the underlying framework as a separate product once it worked.
The hardest part of a serious internal agent isn't the proof-of-concept prompt, it's assembling the tools, frameworks, and infrastructure to keep improving it over time.
05Lessons from OpenClaw: Soul MD + CPU Use
Give your agent an explicit identity file (a soul.md or instructions.md) instead of relying on the model's default personality; that's what makes an agent feel owned rather than borrowed.
An agent performs better when it has its own dedicated space to work in, mirroring how giving a new hire their own computer is the first real productivity unlock.
Not every agent needs to run continuously: idle agents can go dormant like serverless functions and only spin up compute when there's actual work to do.
06One God Agent OR a Team of Agents
Default to one orchestrating agent per company rather than a scattered set of independent bots; the single agent can still route sensitive requests to permission-gated sub-agents.
Access control is the real design problem, not model choice: decide up front who gets read, write, or admin access to which categories of company information.
07How to Build Your Own Agent: Skills, Tools & Permissions
Think of an agent as a hierarchy of files and folders: an instructions file defines its purpose, a tools folder gives it capabilities, and a skills folder encodes how it should behave.
When an agent's output goes wrong, the fix is to update its skill file with the specific standard you wanted, not to individually correct or scold whoever used the agent.
Decide explicitly whether junior team members can let the agent take direct action (like publishing to WordPress) or whether it should draft and hand off to an authenticated human.
08Proactive Agents, Triggers & Self-Improvement
Wire agents to real business events, a failed payment, a new email, instead of only responding when someone remembers to prompt them; that's where most unrealized automation lives.
Build a feedback loop into every agent interaction (a simple thumbs up/down works) and have a recurring job turn the negative feedback into concrete self-improvement changes.
Agent personality is tunable like any other spec: if a team says an agent is too verbose, that's a fixable trait, not a fixed limitation of the model.
09Who Will Build Agents & Getting Started
Pick one boring, well-defined toil task you'd automate immediately if you could, and build the agent's first skill around that, rather than trying to cover the whole business at once.
Non-technical people can and should contribute directly to an agent's skills and knowledge; the deeper technical work is governance, data access, and security architecture, not prompt-writing.
A useful test for whether an agent skill is ready to ship: can you describe the benefit, the cost, and how to get it in one simple sentence?
10The Future: Cheaper, Faster Models & Open Weights
Expect the cost of intelligence to keep dropping and speed to keep improving quickly; open-weight models are already shipping faster variants within days of release due to inference-provider competition.
Match model choice to the interaction type: fast, cheaper models for interactive back-and-forth, and slower, more expensive reasoning models for overnight jobs where nobody is waiting.
Open-weight models can already beat frontier closed models on specific narrow tasks, like a cybersecurity eval, so it's worth testing multiple models per task rather than defaulting to one provider.
11Closing Thoughts: Owning Your Own Intelligence
Businesses are increasingly wary of wiring a single AI vendor's chatbot directly into their workflow because it creates lock-in; owning your own agent avoids that dependency.
Treat your company's agent as being as foundational to your identity as your domain name, something you own and shape over time, not something you rent from one provider.
Glossary
Terms worth knowing.
Eve
An open framework built by Vercel for creating and deploying custom AI agents, structured as a folder of instructions, tools, and skills files.
soul.md / instructions.md
The markdown file that defines an agent's identity, principles, and purpose, functioning as its foundational document rather than generic borrowed intelligence from the underlying model.
V
Vercel's internal company agent, accessible to nearly the whole company through Slack, that routes requests to specialized sub-agents and internal tools.
d0 (D-zero)
Vercel's internal data-analysis sub-agent, connected to the company's data warehouse, that answers business-intelligence questions about customers and product usage.
AI Gateway
A Vercel service that lets developers and agents pull inference tokens from any model provider through one interface instead of integrating each provider separately.
Vercel Connect
A Vercel capability that securely connects agents to over 100 external systems, like Stripe or email, with developer-controlled access rather than blanket read/write permissions.
serverless agent
An agent architecture where compute only runs when the agent has active work, going idle like a hibernating server rather than running continuously.
god agent
A single orchestrating agent that routes every request to the right specialized sub-agent or tool, as opposed to a company deploying many separate, disconnected agents.
“What I believe will happen in the future is that even before you build a website, you're gonna build that agent that's gonna help you build a company.”
Single declarative sentence that reframes the entire episode's thesis, no setup needed.→ TikTok hook↗ Tweet quote
31:40
“You work on the content writing skill of your Eve agent... You're putting that intelligence into the agent in the form of skills, in the form of tools.”
Concrete before/after story about a bad blog post that lands a reusable management principle.→ IG reel cold open↗ Tweet quote
53:20
“It's not about having Claude in your workspace. It's about having an intelligence of your own.”
Sharp vendor-lock-in soundbite that closes the episode's argument in one line.→ newsletter pull-quote↗ Tweet quote
35:00
“Most of the world still thinks about agents as something you prompt. I think there's a lot of alpha in thinking about, can we sort of automate even the prompting such that the agent can be doing useful work for me while I'm not in the computer?”
Reframes the whole 'agent' category in one breath, works as a standalone hook.→ TikTok hook↗ Tweet quote
15:00
“Claude is Anthropic's agent... it's not truly yours. It doesn't have a soul of its own.”
Names the exact reason off-the-shelf chatbots feel disposable to a business owner.→ IG reel cold open↗ Tweet quote
48:20
“My biggest problem isn't like, oh, I wish this was better. It's just like, why did I have to wait fourteen minutes for this?”
Relatable complaint about AI speed that sets up the guest's speed-focused predictions.→ TikTok hook↗ Tweet quote
Topic Map
Where the conversation goes.
00:00 – 07:26denseAgent adoption landscape and killer apps
07:26 – 14:32denseVercel's internal agent V and sub-agents
14:32 – 22:20denseOpenClaw lessons: soul.md and compute access
22:20 – 27:11steadyGod agent vs team of agents and access control
27:11 – 33:52denseBuilding an Eve agent: skills, tools, permissions
33:52 – 40:30steadyTriggers, events, and self-improvement loops
40:30 – 46:22steadyWho builds agents and how to start
46:22 – 53:05denseModel speed, cost, and open-weight competition
53:05 – 54:22sparseVendor lock-in and owning your intelligence
The Script
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
17px
metaphoranalogystory
Really important thing about OpenClaw, which is soul.md. So it's like this soul of your agent that's gonna help you run your company, for example. What I believe will happen in the future is that even before you build a website, you're gonna build that agent that's gonna help you build a company.
Most of the world still thinks about agents as something you prompt. Can we sort of automate even the prompting such that the agent can be doing useful work for me while I'm not in the computer?
Today, I'm having a conversation with Guillermo Rausch, the CEO of a multibillion dollar company, Vercel. And today, we're talking about agents, specifically how companies are using agents within their business.
In this video, we talk about Vercel's internal agent that almost 1,000 people use within the company. We also talk about whether companies need one god agent or a team of many agents.
We also talk about the challenges of setting up agents right now and how to get started building agents that actually improve your business workflows. We also talk about open source models like Kimi k three and a lot more. My goal with this conversation is to answer the following question.
How do we, as business operators, employees, and individuals, use AI agents to be more productive. And if you like videos like these, please consider hitting that like button and subscribing to this podcast.
It helps me out a ton. Let's dive in.
Guillermo, thank you so much for joining me on this on this episode of Agent Native. It's great to be here.
My first question to you is, you know, obviously, we have all these models coming out. Right? You we have Kimi, models from China, models built in The US, Claude, Fable, now Claude Opus five.
We have all these different platforms people can use. And my audience are most people are business operators.
They work in a big company. They want to use agents in their business to become more efficient and to become, like, a better team.
Where are companies at in terms of implementing AI agents in their business? Yeah.
When I think about
we can call it the agentic revolution. Just like any new platform that has hit the Internet or the software landscape, you think about the killer apps.
Right? When the personal computer came out, you know, what were the killer apps? The word processor.
You know? For some of us, video games on our personal computers and things like that. Then mobile came along.
Right? And I think the killer app of mobile in many ways was, you know, not only shrinking interfaces from things that we used to use and putting them in a smaller screen, but enabling entire new use cases.
And I think with agents, see a similar thing. So number one, clearly one of the killer apps of agents is building software.
In building software or, you know, what you could call coding agents, happens to be a core capability of solving a number of knowledge worker tasks. Because when you think about, okay, I'm I'm preparing a presentation for somebody.
You occasionally will say, well, we have to do some data science over here in order to then, you know, get a report or get some data back and put it into a slide. Or you'll say, I'll automate a bunch of different steps and summarize some documents, and then I'll put some other information into a slide. And and so I think clearly one of the foundational parts of this, uh, new period of time where we do a lot of our work increasingly with agencies, coding as a capability.
I think that's this has transformed everyone's jobs. Right?
You can think of it as a number of sort of levels of expertise, I guess, when it comes to coding. So there are people like myself that can do agentic engineering. Meaning, you know, I've been programming for twenty years and now if I sit down and and face a really hard engineering task, I will use a coding agent to enhance my engineering.
Then there's this new emergence of what you would call vibe coding. Right? Which is everybody building a prototype of software or even a full stack application, depending on sort of where your ambitions are and maybe even how ambitious the application itself is.
And so you have products like v zero and Lovable and things like this that are making it more I guess they're democratizing building software or even building the the creative act or enhancing the creative act of coming up with new software.
I also think agents are one of the killer apps is what I would call the run your company better agent or the knowledge base plus data analysis plus project management agent.
The the sort of brain agent that sits alongside of you and disseminates knowledge, business intelligence.
Even day to day tasks like, you know, who should I talk to within the company that is an expert in a certain task? Like navigating the org chart, navigating the what is too many overwhelming amounts of information that reside in the internal systems, uh, of a company.
So to think of this as like making the company's back end more efficient. And as I mentioned, I think coding is this omnipresent capability.
So to give you a concrete example from within Vercel, what we noticed pretty quickly is that anybody that's helping a customer, anybody that's trying to close a sale, anybody that's even building new software, needs to needs to ask questions about, you know, what are our customers doing?
When did they first reach out? How much time do we spend with them?
How much do they use our platform? How many SQS of our of Vercel does this customer use? And so this internal brain agent has sort of emerged as, I think, one of the killer apps of AI.
And maybe for a lot of people, this still seems foreign. Like, what are you talking about? There's an agent that can run my company?
So excited to make that more of a thing. It all sounds like amazing in theory. Right?
Like, the this brain agent that everyone at a company can talk to, it kind of understands kind of the SOPs and the rules of the company, the best practices, that type of thing. And I've been trying to implement this.
You know, I have a nine person marketing team now that, like, helps me create content on my channels, on other channels. And my question to you is, like, you know, for me, when I use AI personally, I'm inside Codex.
That's just the tool that I've been using because I think it's good for knowledge work because I can ask it to create basically any type of document or something, and it'll kind of open up in the side window. But what I can't figure out personally is, like, how do I enable this for a team? You know, if I were to onboard someone new and I want them to have access to my skills and but also, like, a lot of my skills involve my personal connections, like my personal email.
So they can't actually get access to that skill because there's all these, like, permissions that I need to keep separate. But then at oftentimes, I want them to be able to use the same skills that I can. And so I'm wondering, like, at Vercel, like, are you guys kind of trying to create this internally?
And, like, how do you get across these barriers? And, like, what is the actual interface of using agents within a team? Yeah.
Even if you have a team of 10 people or a team of hundreds of people like at Vercel,
I think the way that I think about tools like Codex is that or ChatGPT is that they give you a taste of what AI can do. But your job the new job of someone that runs a company is to actually enable their workforce with agents and to work on the agent.
I think the future of what you would consider to be your intellectual property of the company or your edge against competitors is the ability to create, tune, optimize, and disseminate these agents internally.
And and and, you know, while you can have this sort of a moment when you use something like ChatGPT, maybe to give you an example of our internal agent is called Vee. So anyone within Vercel can go into our Slack workspace and say at v and sort of navigate their day to day.
Whether it's you you give a great example. So if I need to create new content. For example, our marketing team needs to help promote a new product that we worked on Or communicate a product change.
Or write an engineering blog post in in collaboration with an engineer that work on a certain capability. All of this goes through this v agent. And this v agent has a number of skills that we continuously sort of update and improve.
It has sub agents. It has sort of imagine the ability to create like a virtual employee team.
So there is the content agent that is really good at writing marketing materials. There is the data an analysis agent.
We we internally call this d zero, but it's one of the it sort of think of it as, like, the the nexus of intelligence within our company. Like, anytime when you get information about how a customer is doing or, you know, how they could use more of our cell or things like this, we have this sort of d zero agent that is connected to our data warehouse.
And so the experience of using an agent actually ends up being extremely user friendly. Why?
Because all you need to do is you join Vercel, you join our chat workspace, and now you sort of have this omnipresent intelligence that can help you.
And now you might, you know, you might go to Vee and say, hey. Can you change some information on the website?
And so Vee can still sort of coordinate with other agents. It could it could delegate a a task to Codex if it wanted to. If it it can create a prototype with v zero.
It can query Vercel to get information about our production systems.
But I think what's what's key is enabling every company in the world to sort of deploy this brain and this intelligence and continue to sort of optimize it over time. I have a lot of questions based on this. My first one is do you have, like, a team that manages v?
Like, that where okay. You have a team. What what does that team look like?
You know, how big is it, and, like, what do they do on a day to day basis?
So maybe to back up, I wanted to share a little bit about our product development philosophy at Vercel. When we have a vision of the future that can be informed by, you know, pains that our customers have or things that we notice internally could be better, we try to solve that problem ourselves first.
So this idea of let's have an agent that can help with every aspect of our job sort of emerged pretty obviously. Like you mentioned, like, anyone that uses ChatGPT notices, oh, it can reason.
But ChatGPT doesn't have access to my internal knowledge base and customer records and the set of best practices of how we build software, etcetera. And so the inspiration was anytime you talk to somebody, could there have been an agentic intelligence layer that could have gotten you that information sooner?
So that was sort of like the inkling, the inspiration for it. Next thing is, how do we build this? And so Vercel has built a number of agentic infrastructure services and tools.
Right? So we built the AI SDK that helps developers talk to any model in the world.
We built AI gateway, which helps you get tokens from any model in the world.
At the end of the day, you know, what we realize is that, okay, if there's an agent like V, I don't want it to necessarily be Claude or Codex or OpenWeight. At the end of the day, the customer doesn't matter.
And ideally, we autonomously choose the best model for each task. So we almost thought of v as a superset of all agents in the world.
And so we designated a few folks to sort of like try it out and build this conversational experience. First, it started out as a support assistant. And that alone was extremely useful.
Why? Because we are hiring new people. And also in Slack, we talk to a lot of our customers.
And so anytime that you have a question about how Vercel works, we wanted to have an at Vercel functionality that could know anything about Vercel.
And that in itself was super, super, super helpful because it became sort of like this easy way of giving support to our customers. But the difference between an AI assistant and an agent is that an agent can do things for you.
And so we started thinking in terms of skills and in terms of jobs to be done. So this we gave it a name, so v for internal purposes. And so we wanted to have a clear distinction between the customer facing agent, the agent we give users of Vercel, which is Ad Vercel, and the agent that runs our company.
So v is sort of the shorthand for this. So we created the v team. The other thing we realized in this process, and maybe this goes at the heart of your question, is it's actually pretty hard to assemble all of the tools, all of the frameworks, and all of the infrastructure to make something like this happen and to improve it over time.
And so that gave inspiration for us to we built Vee and then we shared the framework that we used to build it back to the world. We call this Eve.
You might you might sound like we're super creative with our names, V, Eve, Vercel. But Eve is sort of the, you know, Next.
Js or React, what they did to the web, they made it really easy to build websites and web applications. The thing that I think every knowledge worker, every individual, every entrepreneur will want in the future is to have an agent that they can call their own.
And this is what we're helping people enable with with Eve. Got you. Yeah.
I think, you know, this is something I've spent a lot of time thinking about. Like, how do you give the normal person the access to not only just have an agent that has a bunch of context, but to also kind of, like, customize it.
And I think although it feels like now that OpenClaw was kind of a fad, you know you know, if if you look at the Google Trends, it's, like, gone way down. I do think it unlocked kind of a magic moment or I there's a reason it went viral in the first place.
It wasn't because there was some secret paid promos by Openclaw. I think there was a genuine desire for people to put an agent on a computer and let it do things for you.
I had a lot of epiphanies,
uh, from OpenClaw that informed the development of Eve. I think you're absolutely spot on. One of those things is that OpenClaw showed just how much a coding agent can do back to my, uh, initial point.
Like, what is OpenClaw fundamentally? It's the raw intelligence of the model plus every tool at its disposal.
Right? Like A full full access. Yeah.
It can write code, it can run it, and it can have access to everything.
And that's magic. To the point where it could do things accidentally. And, like, I think that's I remember listening to Peter who created Openclaw.
He said something like that he, like, asked for something and then it, like, gave it found an API key on his computer and it did something that he didn't even ask for. And I think that was kind of the magic moment.
And, you know, they added, like, the heartbeat, which was this thing that kind of like initiated it to wake really important thing about OpenClaw,
which is soul.md. So when you when you create an OpenClaw or when you use OpenClaw, you're not just taking the off the shelf agent that somebody else built.
Clearly, Claude, for example, it's a great agent, but Claude is Anthropics agent. It has its own set of principles and and and sure they will they they give you ways to customize it and whatnot, but it's not truly yours.
It doesn't have a a soul of its own. Right?
And so I think that was another really big unlock, which is what is the soul dot m d file? It's just it's just literally markdown text that defines the genesis of that model.
So when you create an agent with Eve, which we basically learn from that and and basically an Eve agent at its most basic is a folder with an instructions dot m d file in it.
So it's like the soul of your agent that's gonna help you run your company, for example. And then the other thing that we learned is it's awesome that it can run code, write code, it has a computer for it.
Right? Like the the whole like Mac mini thing was actually quite meaningful.
Right? Like people realized, okay, this agent can do anything under the sun, but it's dangerous.
And it needs a space. It needs its own like thing. It is it it give the agent some space.
Right? Like and so people bought Mac minis. And and and that basically giving an agent a computer massively improves its performance.
Its reasoning performance and its ability to deliver outcomes for you. And so what's really fascinating is it's not too unlike hiring a knowledge worker.
What is the first thing a modern firms does when they hire a human? Here's your computer.
IT gave us a MacBook. It has a bunch of programs installed. It's logged into all of your key systems.
And so we wanted to give you that as well for your own agents that you build, but we wanted to build a secure and efficient environment for you to run.
And so the security part is that you define the tools, the human in the loop approvals, and the data access controls for anything that the agent can do. And the other aspect of it is it doesn't assume that the agent is always running in a computer, which is actually kind of counterintuitive. I just said an agent gets better if he has a computer, but not every agent needs a computer that's running twenty four seven.
And so in in in our in our lingo of the Vercel and and cloud world, we call this serverless. The idea is that if the agent is not doing anything, it can go to sleep. Maybe another metaphor is imagining a Mac mini that hibernates when the agent doesn't have anything to do so that it doesn't use electricity.
And so because we at Vercel, we run, you know, billions of deployments, we needed a mechanism such that agents can be very, very, very efficiently operated and run. And so that's another sort of ingredient that we learn from the the OpenCLOS of the world.
Okay. If we're gonna run these things at massive scale and we need to run them securely,
how can we create infrastructure that enables that? Got you. That makes sense.
Yeah. I think I think all of the the big AI labs who've re who've who are, like, kind of releasing a product that is an agent on a computer is trying to shake it into people. They're like, this is a computer.
It has a computer, and it's not easy to communicate to the average people. You know, OpenAI is struggling with that right now where they're, like, literally tweeting.
They're like, GPT Work is an agent with a computer, and it's not easy to convey that as you interact with a chatbot.
You know? Like, it's like, what does that even mean? And, you know, and and I'm even I'm even struggling with it, you know, and I think, you know, and I think you can kind of divide whether you look at Anthropic or OpenAI, like, you can kind of divide their products into, like, how much computer access they have.
It's like the chatbot doesn't have any computer. GPT work has some computer. It doesn't it can't run terminal commands, but then codecs can run terminal commands, but you can only get it on your computer because they don't have and so I think that is actually the computer aspect of agents, I think, one of the parts that makes it really confusing at this stage right now.
I agree. And and my goal with
the agents that we build in in Vee is that, you know, whether you're an intern that just joined Vercel or you're a super experienced engineer or you're somewhere in between, I don't think whether I I think that's sort of the implementation detail that the agent builder needs to know about.
You need to what I want for the future is that someone that's building an agent can very carefully define governance, data access control in the security model.
Right? For because agents are interacting with customer data. So you can't just be like, I don't know, man.
It rests a computer and has access to like all of the databases of everything. You have to be really, really, really thoughtful about it. That's literally our new job.
Right? And but whether it runs one or it runs a million computers, completely inconsequential to the end user.
In fact, you know, you can think of this agents as being orchestrators.
In fact, when when someone goes towards Slack and says at v, they're really talking to the orchestrating agent.
The one that could delegate a task to a million computers, to one computer, maybe even no computer. You know, we have customers of Vercel that have built agents that have so much usage that they figured out ways to make the computer smaller and smaller and smaller just for the sake of cost efficiency.
And so I I think the my hope for the future is that the very technical people can sort of know, like, oh, this particular conversation with this agent resulted in all of this usage of computers and whatnot. But for the most part, it's all about getting high quality outcomes, high quality analysis, high quality, you know, accurate information.
Performance is becoming more and more of a the dog of town. Right? Like, people really care for fast models and fast execution.
So that's another aspect of, like, how do you get your agent to be delightful.
Okay. So let let's say for a sec I wanted to create a v agent for my team. Yeah.
My first question with this, and and this is something that I've realized talking to a lot of business owners who are, kind of know about agents and they're they're trying they they're they're confused on whether you want one agent that's like a god agent that knows everything. Or if you want a team of agents that sort of, like, share a knowledge base.
Because the conversation that I'm having with a lot of business owners is like, well, the marketing team has access to these things. And the finance team, like, I don't even I I don't even want the marketing team to know about certain finance documents. Totally.
And so, like, that's my question is, like, how if I were to be creating my own VAgent for my company, how do I think about that?
You know, or yeah.
So first of all, I'm a user experience guy. You know, I started Vercel because I was frustrated with how slow creating software was and how slow the average website and web application experience was.
So I always try to work backwards on the user experience. The ideal user experience with an agent is the Star Trek computer or the Ironman JARVIS.
It's ambient computing, and I don't need to target a specific capability. That's why we are we're reasoning with agents to begin with.
It's like there's probably, like, hundreds, if not thousands of internal tools that people at Vercel have built that I don't even know they exist, frankly.
There's just too much. Right? And so when you have this as intelligent agents, they can act as routers.
V, our internal EVE agent is a router. So if you ask it about Vercel knowledge, it goes to the capability that we have for looking up our documentation or knowledge base, etcetera.
If you ask about if you need to help a customer with a support case, it has a support agent within it that has access to our support ticket infrastructure. Okay.
So that that answers sort of my perspective is that it's more on the god model. And maybe to give you a metaphor, because I really think that what we're doing here is we're redefining how companies of the future will work.
When you join a corporation, they might give you a corporate phone. And that corporate phone is already preconfigured with your identity and with a set of applications.
You have the application for the, I don't know, internal chat. You have the application for this and that. So I think the internal agent that helps you run the company is not not unlike that.
It's the job of the new sort of IT department is to say, what are the capabilities that we're bundling into this agent?
And also, crucially, how do we manage identity and who gets to access what information? Which is also extremely business specific.
It depends on how regulated your business is. If you're a small startup, I can believe that, you know, your nine person team, they all have pretty equal access to most of the information of the company.
Maybe two have information to the financials or or maybe the decision I remember when I started Vercel was like some of us had, you know, read write admin. And but I think most of the first 10 person Vercel team had read access to almost everything. Right?
And so the job of the person that works on this foundational agent is to determine the the access control, the tools, the guardrails, the the audit trails.
And and and like I said, this is actually pretty hard work to do And and and why we wanted to create a framework that made that the fundamental job because, you know, wiring up the model,
wiring up the infrastructure, all of that, we can sort of customers can offload to us. That makes sense. And so I yeah.
I guess the agent would also be able to see where the message is coming from. So it's like, okay. If it gets sent in this channel, it'll delegate to this sub agent or access these certain files.
That makes a lot of sense. Totally. I just get I guess, because what you're telling me is, like, so appealing, like, being able to create your team's agent.
And I don't think anyone's cracked the interface for this yet. And I know you guys are building a framework. You deal with a lot of developers.
I guess what I'm dying for is, like, a way some sort of interface to understand it.
Because even the technical people like, I've even showed technical people, Eve, where I like, I'm like, can you help me make sense of this? And I think it's still at a stage where it's not super easy to, like, fully understand. And so I guess yeah.
I just wish there was, an interface where I could go in and, like, set these rules. Maybe I'm talking to an AI and it's configuring it. I I guess I The the way that most of these agents are built is that you're talking to an AI
that is helping you maintain your Eve project. You'll hear me use the word file system or folder a lot.
I find that it's it makes the world really easy to understand if you think it if you think about it as a hierarchy of files and folders. So the way that a Neve agent works is that you started with that instructions file that says, you are the agent that helps run Riley's business.
You can even have some context about who you are. Like, our business is in you know, we disseminate information about AI and our values are transparency, we're not opinionated, and we love shipping things.
Like, something like that. Right? Okay.
But that agent still knows nothing. It's a tabula rasa. It just has the raw intelligence that comes from the model, and it has a a basic set of instructions.
How can he do something useful for you? Well, you talked about, okay, let's help the marketing team create content.
And let's say that one of the things that you really care about is posting on your blog. Okay.
So in an Eve agent, the first thing you do is you can create a tools folder and you can always start exposing tools to the agent. And so you can say, let's say that your blog is running WordPress or some system like that.
Now you can say to the agent, now you have a tool to read and write blog posts to WordPress. Okay.
Great. You you created that file, wordpress dot t s on on that folder, and then you ship your agent.
You you use the word channel, also very important. The this agent needs to communicate to your team in some channel. So Eve supports every channel under the sun.
It can be WhatsApp. It can be Telegram. It can be Slack.
It can be Microsoft. IMessage? It can be iMessage.
Yes. Amazing. And so the next question is, okay.
I created the agent. I gave it this sort of soul. I gave it access to WordPress.
Now, you hire an intern. Can the intern ship any blog post that it authors together with your internal agent to prod?
You probably don't want that. And so this is the job of like at at some point, maybe Riley, you are working on your ev agent or someone in your team you designated as sort of the agent administrator. You're gonna say, okay.
If the person lives within a certain part of the organization, we let them write directly to WordPress.
Another approach that I've seen people take is that when they interact with the intern over Slack or over Telegram or whatever, you have to authenticate with WordPress. So you delegate to an existing permission system that you already have.
So the Eve agent ends up being sort of the facilitator of the transaction, but it doesn't have direct access to WordPress itself. It it it will help you sort of draft up the content.
So this is just an idea that we cooked up in this conversation. But imagine that every day you start realizing, that's really powerful. I just unblocked my entire team to be able to draft up blog posts that go directly to WordPress.
But next time, tomorrow you hear an escalation and you hear, hey, at Riley, I just saw your blog post. It's I read your most recent blog post. It reads like complete clot slop.
What do you do? And you go you go into your team and say, guys, what did we just do? We became really productive and we started shipping a lot of slop.
You know what you do next? You work on the content writing skill of your EVE agent.
And so this is the meta work that we will all be doing in the future. We're not working on the blog post itself. You did not go to the intern and scolded him for like, hey, what what do you do?
You shipped a bunch of slob. You're putting that intelligence into the agent in the form of skills, in the form of tools.
Uh, and, uh, of course, over time you can get more sophisticated and and, uh, it's not just about blog, like how can we infuse the content writing capability with
what people are saying on x about your business? I was gonna say that, like, a lot of the skills that I find very useful for content ends up just being, like, grounding in some relevant source.
And so you can put I call them, like, plugins, like, where like, there's one called scrape creators. It's some API that I found that scrapes content from certain channels. And so, like, before it ever writes anything or before it ever ideates an idea for YouTube or a or a packaging concept like a title and thumbnail, it'll go and, like, look on social media and find those things.
Totally. Yeah. And that's another thing.
Like, okay. So if I'm creating a v agent, yeah, I'd I'd want to add certain APIs, and you can add I I would yeah.
Call you can call them plugins. Or, like, how do we distinguish between plugins and skills? Can you add plugins to skills, or how does or are they all just skills?
Text.
So going back to you you got that escalation that says, Riley, you just shipped you're shipping a lot of blog posts, but they all they have too many EmDashes. And so this is what's beautiful about that idea of it's just a folder.
You go into your Eve agent and in the folder skills, you say content writing dot m d and you say, this is how we write. This is what I like.
This is what I don't like. You also talked about, I I think that the future of work will be the agent becoming a lot more proactive as well.
So Eve can have a schedule, for example. Every day at night, it reads social media.
It parses keywords. It gets replies from your posts. And from that, it can do something.
It can draft up new content. It can even give you a report inside of Slack. And this we actually have found to be extremely helpful out of our self.
The idea that our agents proactively give us information. So every Monday, I have a I have my internal agent give me a download of what's happening across every product area.
What are the key metrics that I care about? So you can have the agent be doing thinking in the background on your behalf. And I think it's not just about I think most of the world still thinks about agents as something you prompt.
But I think there's a lot of alpha in thinking about, can we sort of automate even the prompting such that the agent can be doing useful work for me while I'm not in the computer?
I think one of the limitations for me and
I've been able I I have a lot of automation set up that trigger an agent to do a certain task, and it is really useful. One thing that I'm struggling figuring out how to set up, especially at my at the team level, is to get outside things to trigger the agent, you know. And there's many ways I think you could do this.
But, yeah, like, do you guys have any of of that setup? Like, if some event happens Totally. It automatically okay.
Yeah. Can you talk about So
I think events that originate in systems like Stripe, like there is a refund request.
We make it really easy to connect all those systems. In fact, when we sat down and we thought about what makes it really hard to build an agent, It's actually not the proof of concept part because anybody in the world can sit down, open Clot Coder Codex and build an agent.
The sense that like when you're prompting it, you realize what it becomes capable of. What we talked about with Open Clot, the raw intelligence is already there.
What's hard is securely connecting it to your systems. So we built a capability on Vercel called Vercel Connect that gives your agents access to a 100 plus systems, but it doesn't just give them full read, write everything access right away.
It gives you the developer the control. And that might mean that you subscribe to an event and then you send it to your agent. You can say, hey, every time Stripe has a failed payment, let the agent know.
Every time we get an email, let the agent know. And so you start thinking about the world in terms of events.
In fact, I mentioned that a lot of our agent interactions are happening on in Slack. Slack is just another event.
It's someone said something and the agent that gets fed into the agent's brain. And so any any connector of this sort of
repertoire of connectors can originate some kind of behavior in the agent. Got you. That makes sense.
Yeah. That's just something we've been thinking about a lot because you're right. Everything is just an event.
It's just things happening. And then when something happens, if an agent can take care of it, they it should take care of it.
And I think I'm like I've automated none of that in terms of what I could possibly automate, which is really interesting. I'm mental model. So I mentioned that the the thing that I'm excited about with Yves is that when when I started Vercel,
the most imminent thing that I needed to build was a website. Like, it felt like, how do I put my fingerprint in the world?
What is one of the earliest things that you do when you create a company? You register in Delaware. If you are in The United States or even internationally, you incorporate, you choose a name, and so you register the domain name and you ship a website.
Even a website that says like, hey, we're in business or welcome to the minimum viable sort of identity of your company in the Internet. What I believe will happen in the future is that even before you build a website, you're gonna build that agent that's gonna help you build a company.
The it's gonna be your factory. It's gonna be the the trusted partner and adviser in everything you do that's constantly learning about the trajectory of your business.
And so it's extremely critical that as you sort of evolve your business, this agent gets access to more of these data streams of knowledge and information.
And everything really is an event in this world. Another important factor there is self improvement. So when whenever you start a company, you're constantly learning.
You're you're teaching your employees. You're helping them, you know, learn from mistakes, learn from incidents, learn from customer feedback, etcetera. It's gonna be very important that your agent over time can improve.
And so with Eve, we thought about, okay, if there is a baseline of information that your agent has, how do you evaluate the agent?
Can you write tests or can you give it exams so that you actually know that you're making forward progress as you as this agent sort of gets more sophisticated and more capable over time?
And so think of this as sort of even more fundamental than
the .com of your of your of your company. Yeah. And do do you guys, like, put evals into Slack?
Are there any ways to, like, evaluate whether an agent does well or doesn't do well? Like, could you based on someone like, could a employee who got a response from V, could they say, like, oh, this wasn't a good response and okay.
They can do that.
Yeah. So the every response that we give on Slack has a and and by the way, maybe to also give kudos to the Slack team, like, Slack is kinda becoming like an agent operating system of sorts.
Right? Because, like, it used to be for messages between humans. Now it's humans and agents.
And so they have built UI that is just really easy for the developer to add. Right?
So, like, the thumbs up, thumbs down thing is super easy to add. And so every Eve agent that we create, for example, at night, we can have a job that aggregates all of the negative feedback and proposes the next stage of self improvement.
We can say, hey. We got five thumbs down on these answers. What are the things that the agent itself can even propose how to improve itself?
Oh, I missed this. Oh, this person critiqued this part of my response, or they said I hallucinated or whatnot.
I do think it's very important that humans are still involved in that loop. But I think increasingly more and more of the job of get the agent getting better is also being done by the framework. So the framework itself comes with evals that, you know, are basically test cases.
Right? When you build a web application or a website, you write unit tests and you make sure that the logic is sound. When you create an Eve agent, you write evals also to ascertain that the logic is sound, but that the information it gathers is is sound and and it's accurate.
There can be evals about personality. At some point, we're hearing from people that our internal company agent was too verbose. It was speaking too much.
Uh, and so you we kinda basically gave it a better personality, and and you can create evals around that as well. So do you do you view this, like, the near future, like, over the next few years? Do you think it's just gonna be mostly technical people building agents for companies, or do you view this as something that whether you can code or not, you'll you'll be able to create agents for your team?
So
because building software is being so democratized, um, think of it as, like, again let's go back to that idea of, like, I'm starting a company and, like, the first website that I built is sort of, like, I could have used any service on the planet.
Drag and drop, give me a free website with my domain name, like, anything like that. So I think that first building block of your agent, everybody's gonna be able to to create.
I think over time, I mean, the whole business runs on this. Hundreds of millions of dollars of revenue are dependent on the well-being of this agent because our sales reps depend on it.
Our support team depends on it. I depend on it. And so you this is a very important piece of software.
And so I think it's a combination of everyone can contribute to the agent, information, skills, critique, feedback.
And then there is engineers that are working on the core system loop, the access to data, the governance, security, all of those pieces that I think need to be more technically minded.
But I don't think that the code writing part is as important these days. It's I think I would describe it as people that really understand data flows, threat models, and architecture of systems design so that they can, like, carefully think about the the, again, the operational excellence of the agent and the security model of the agent.
Very interesting. Yeah. Um, because, yeah, I think there's a lot of people, business owners, not all of them are technical, who are reaching out and they're trying to create agents.
And so I'm just trying to, like, leave people with, like, a a tangible thing that they can do, like, a point to a place where they can go to kind of build their first agent or build their v. Because I think with what I've realized with these agent tools, all of them, is I we we we can have conversations about it.
We can talk about it. I can learn. I can use AI to, like, learn about it, but nothing hits like doing it.
And I think that's kind like like, once you do it, then you're like, oh, I can do that. That means I could do this thing, this thing, and this thing, and, like, kind of your world opens up as you do even the most trivial things. And so, yeah, I A combination there would be, you know, what I've seen give you an moment
is create an eve agent. Go to eve.dev, deploy your first agent, but connect it to your favorite chat medium.
If your if your company works in Slack, connected to Slack. If you like WhatsApp, connected to WhatsApp. And pick one boring or, you know, kind of pick a toil task of your business that has a system to it, but it's not you know, it's something that if you could automate it away, you'd absolutely automate it away and write down the scale of that task.
It could be, for example, something we do a lot at Vercel is we put a lot of work into drafting up our product change log.
When you go to vercel.com and /changelog, Every piece of content there narrates the storytelling or evolution of our product. And in many ways that change log is a grounding for my engineering team.
How do I know if an engineer is being productive or not or whatever? Like, well, one of the things that I do is I I measure it by have you shipped something that we can communicate to customers is an improvement to our platform?
So one change log that's about to go out, maybe by the time you watch this has already gone out, is we we improved the end to end deployment process of an application or agent to Vercel by seven seconds. Seven seconds we've shaved off over a lot of infrastructure work.
So when you go to vercel.com/changeover, you're gonna find that. We improved our product and we shaved down seven seconds.
So it used to actually take a lot of work for an engineer that is in the depths of infrastructure to collaborate with the marketing team and get that thing out into the world. Because we have an agent internally, we've cut down that process into one Slack thread that the engineer creates.
The agent refines what they're telling me because, you know, engineers are sometimes so in the weeds that they struggle to communicate things in a way that is I call it context free.
You know, maybe they start talking about, you know, computer science or like I'm just hey. Can we boil it down to the business benefit? Simple.
Seven seconds. Uh, it's enabled for every customer. It's free.
So that's kinda like a little formula that I have. People wanna know what's the benefit, how much does it cost, and what do I do to get it.
Mhmm. And so that formula that I developed over many years of product marketing skill, I put into that EVE agent.
And so for the listeners, think about something like that. Maybe it's, like, quote, unquote, a secret sauce of something you do really well, but takes a lot of time, and you wanna do more of it.
And so start with that skill, connect it to a communication channel,
ship it on Purcell. Got you. Okay.
That makes sense. Yeah. I think to kind of I know we're we're running up on our time here, but what are you most excited about in it could be a model.
It could be computer use or some browser use. Like, what unlock do you think we're gonna get in the next, like, three to six months that will make using agents way more fun or way more effective?
Very simple. Cost of intelligence continuing to go down. More intelligence to for more people, more variety of models.
One of the great things about building with Eve and building in Vercel generally is that we give you access to every provider of models and every model in the world. It's model agnostic.
Yeah. Totally model agnostic. Right?
And that plays into your benefit because you retain ownership of your data, of your skills, you get to choose models, and you get to benefit from the competition.
There is some news that's gonna go out tomorrow about models getting dramatically cheaper. Mhmm. Literally tomorrow.
Tomorrow. And if you were building in this way, you're gonna benefit.
So the other one is fast. Models are gonna get way faster. I think we're gonna start seeing what happened with the personal computing and mobile computing revolution, which is that, you know, we got the iPhone.
If you were if you could travel back in time and or even pulled out the first iPhone out of a drawer, you'd be astonished at how slow it was.
The refresh rate. Like, you would open an app, it would do nothing for several seconds, and then slowly at maybe 10 frames per second, the application would show up in front of your eyes.
Right. That's where AI is at today. Yeah.
I think for most knowledge tasks,
like, I just want faster. You know what? My biggest problem isn't like, oh, I wish this was better.
It's just like, why did I have to wait fourteen minutes for this? You know? And it it like, if it was 10 times faster, it be insane.
And I feel like we're pro like, how long do you think it'll take for the models at, like, a 5.6 level, like, so like, soul level?
Days. Days? I mean, maybe days is the most, like, optimistic.
I I think we're literally, like, weeks, single digit months away. One of the data points that I can share is on the open weight, and this is why I'm excited about open weight models.
The competition between the inference providers around open weight is so extreme that GLM dropped. We added it to Vercel AI gateway.
It's an incredibly good model, GLM 5.2. Within days, we had a fast variant that was four times faster.
There we have more providers coming online for GLM that keep raising the bar of token per second performance. Mhmm. GLM 5.2 fast is astonishingly fast.
It's only getting faster.
What did you think of gonna get Kimi k three. Happen to Kimi.
Think we're still in the early innings of that. What did you think of the model, like, in general? Like, do you think it's you think it's really good?
You think it's up to par with, like, an Opus four eight? I think GLM 5.2 is already in that category. I think Kimi raises the bar.
I think Kimi can do things
that perhaps only, you know, fable class models could do.
Not quite in all in all of its dimensions, but for example, when we evaluated it for cybersecurity, it outperformed Opus 4.8 clearly. Um, and it was almost at, uh, you know, sole level, sole still at the frontier.
But, again, this is a beautiful thing about having choice is that depending on what you're doing, you're gonna choose different price performance ratios. Grok for fast and highly accurate.
Like, if I have to choose today a model that's gonna be my workhorse model, that would be like the default. If I have an agent that is my Slack and needs to do a wide variety of tasks and it has to do it quickly because there's another person waiting on their side, I would absolutely go with Grok 4.5 or GLM in terms of, like, price performance.
Mhmm. Now I mentioned the productivity.
What about, for example, at night finding opportunities in our business, crunching data and extracting novel insights for the executive team.
Well, those things I can throw more reasoning power, and it can take more time. Mhmm.
I might even wanna throw a consortium of models at it. Why not have Kimi and Sol and Grok come up with three points of view and then give you the summary?
And this is why I find it so interesting. Right? Like, we're still in the early innings of understanding what are the principles of design and user interface engineering, but for agents.
Yeah. If I'm talking to an agent interactively, I want fast.
If the agent is doing an asynchronous job, I want accuracy. Yeah. You don't care if it takes all night.
Like, it it doesn't make a difference if that yeah. Yeah. That's true.
I I didn't think about that.
We're about to launch a capability in AI gateway, which is you as a developer or even your agent can say, please do inference.
Please, like, get me tokens, but in batch.
And I don't care how long you're gonna take. Mhmm.
Like, you communicate. It's a little bit like putting in a buy order and you're not worried when it gets fulfilled.
Right? Right. Right.
You're just willing to wait. And then anyone in this market can fulfill your order. Mhmm.
Yeah. Almost like a spot market for intelligence. Right?
That makes sense. Yeah. And and this is extremely exciting because you might say, hey.
Like, come up with a proof or disprove of the Jacobian conjecture for two dimensions.
And I don't really care when, but you spend this many tokens. And someone at some point is gonna say, hey. I already paid for the GPU.
It's gonna be the Internet. Yeah. No one is using it.
Let's throw some capacity. It's a little bit like SETI at home
for those who would remember. Yeah. Rent out your spare compute capacity, solve hard problems.
Yeah. Because if you get it next week, it doesn't matter. You know?
You're still solving a really crazy thing. Anyway, I really appreciate you joining.
I think you guys are gonna do great. One thing I didn't realize is how much business owners don't wanna get locked in to a certain provider.
I mean, you know, like, Claude Tag is their kind of I don't wanna say it's their version of v, but it's like kinda an agent you can add to Slack. And so many people are resistant to it because they don't wanna get locked in to only Claude's models.
So I think that is Yeah. Something that you guys will have going for you. That's really cool.
And it goes beyond, you know, the the model. I think it's not about having Claude in your workspace. It's about having an intelligence of your own.
Right. So there's almost like an element of, like, baptizing your agents. Like, this is our agent.
This is our company. It's you know, I actually liken it to the web because the web was all about I own my domain name. Mhmm.
I'm the I'm the king of my own domain. And I think we're now seeing we're living through the version of that for the intelligence age.
100%.
Yeah. I agree. I thank you so much for coming on.
This was this was a lot of fun. Let's do it sometime soon.
Anytime, Riley. Thank you.
The Hook
The bait, then the rug-pull.
Before the interview even starts, the cold open drops the episode's real thesis: the agent needs a soul.md, a file that gives it an identity of its own, not just borrowed intelligence from whichever model is underneath it. From there, Riley Brown sits down with Vercel CEO Guillermo Rauch to work out what that actually looks like inside a real company.
A week where xAI undercuts Anthropic and OpenAI on price, and GrokBot's named-agent structure hints at how every AI platform is about to fix the chat-versus-work problem.
An interview with an early power-user on why the open-source Slack-for-agents app is going viral, then a full walkthrough of running a personal AI agent team inside it.
A first-look test of Codex's new talk-back voice mode: triaging a real inbox, editing a whiteboard, building an iOS app, and controlling it all remotely from a phone.
Claude Desktop and the Codex app both quietly shipped native, multi-tab in-app browsers within days of each other — used right, one prompt now opens, drafts, and stages a dozen tabs at once.
Riley Brown and Ras Mic dig into GPT-5.6, Codex's background computer-use, and why self-scoring agent loops are turning coding tools into a general operating system.
01:26
02:22
03:14
03:43
04:11
05:07
05:46
06:08
07:54
08:29
09:10
09:39
11:04
11:33
12:44
13:07
13:50
14:57
15:49
16:23
17:17
17:40
18:25
19:18
19:45
20:59
21:30
22:26
23:06
23:54
24:29
25:29
25:55
26:34
28:04
29:04
29:29
30:24
30:50
31:26
31:56
32:59
34:06
34:53
35:48
36:13
37:19
37:28
38:03
38:51
39:42
40:54
41:48
42:21
42:55
43:49
44:42
45:27
45:54
47:02
47:49
48:40
49:20
49:39
50:36
51:04
51:41
52:26
53:43
25:21
Video of the Day55:15
16:27
Video of the Day23:10
47:51
21:08