The Future of Code Generation | Guillermo Rauch, CEO of Vercel | Ep. 20
Guillermo Rauch (guest), Jack Altman (host)
In this episode of Uncapped with Jack Altman, featuring Guillermo Rauch and Jack Altman, The Future of Code Generation | Guillermo Rauch, CEO of Vercel | Ep. 20 explores code generation shifts bottleneck from writing code to landing outcomes Rauch traces Vercel’s origins to an obsession with iteration velocity: making deployment feel as instant as editing files locally, then scaling that into a product that ties developer experience to business outcomes.
Code generation shifts bottleneck from writing code to landing outcomes
Rauch traces Vercel’s origins to an obsession with iteration velocity: making deployment feel as instant as editing files locally, then scaling that into a product that ties developer experience to business outcomes.
He argues code generation is rapidly improving, but the core bottleneck is moving from writing code to reviewing, trusting, and “landing” changes in production with measurable impact.
AI is evolving from assistants to outcome-focused agents; this creates psychological benefits and new workflows (e.g., just-in-time internal tools, generative dashboards, live customer prototyping) while raising new risks in fault attribution and security.
Looking forward, he predicts more specialized, vertically integrated agents and an “MCP-like” agentic internet that avoids a single platform gatekeeper, while emphasizing taste, presence, and discipline as enduring human advantages.
Key Takeaways
Iteration velocity is a core competitive advantage.
Rauch’s pre-Vercel lesson was that the most impactful CTO move was making deploys effortless: push to Git, get a live URL. ...
Get the full analysis with uListen AI
Developer experience alone isn’t enough—prove business outcomes.
He frames “landing” as more than shipping code: it’s deployment plus adoption/conversion impact. ...
Get the full analysis with uListen AI
Codegen progress shifts the bottleneck to trust and review.
Even if large portions of code are AI-generated, mature teams can’t merge safely without confidence. ...
Get the full analysis with uListen AI
Vertically integrated, opinionated codegen can beat generic tools on reliability.
Rauch contrasts broad tools (Cursor/Claude Code) with constrained systems like v0 that generate within a known stack (Next. ...
Get the full analysis with uListen AI
Perceived productivity gains often exceed realized gains.
He cites enterprise evaluations where engineers feel dramatically faster, but measured “landed” output doesn’t improve as much—or can worsen—because downstream steps (debugging, review, integration, operational risk) expand or slow down.
Get the full analysis with uListen AI
Agents change software creation—and developer psychology.
As workflows move from request/response assistants to outcome-seeking agents, the emotional experience shifts: the agent absorbs the error-filled struggle. ...
Get the full analysis with uListen AI
The internet remains brittle; security and blame assignment become first-class problems.
He points to supply-chain incidents (compromised packages, Log4Shell) and argues adversarial AI will accelerate attacks. ...
Get the full analysis with uListen AI
Notable Quotes
“To us, it’s not just about writing the code, it’s that you land it.”
— Guillermo Rauch
“The bottleneck has shifted to reviewing that code.”
— Guillermo Rauch
“Agent is the new model.”
— Guillermo Rauch
“It’s a huge house of cards.”
— Guillermo Rauch
“We need to raise the bar where AI is not just a slop-generation machine… think kind of like a Waymo.”
— Guillermo Rauch
Questions Answered in This Episode
When you say “landing,” what metrics do you actually use internally to decide a feature is truly landed (adoption, retention, revenue, latency, error rates)?
Rauch traces Vercel’s origins to an obsession with iteration velocity: making deployment feel as instant as editing files locally, then scaling that into a product that ties developer experience to business outcomes.
Get the full analysis with uListen AI
In the AI-PR review bottleneck, what specific signals would raise “baseline confidence” (formal verification, test generation, runtime replay, policy checks, diff-risk scoring)?
He argues code generation is rapidly improving, but the core bottleneck is moving from writing code to reviewing, trusting, and “landing” changes in production with measurable impact.
Get the full analysis with uListen AI
You suggest a convention for agents to identify themselves to tools—what would that protocol look like in practice, and where should it live (git metadata, IDE, CI, runtime headers)?
AI is evolving from assistants to outcome-focused agents; this creates psychological benefits and new workflows (e. ...
Get the full analysis with uListen AI
Where do you draw the line between opinionated vertical integration (safer) and lock-in (less flexible), especially for enterprises with heterogeneous stacks?
Looking forward, he predicts more specialized, vertically integrated agents and an “MCP-like” agentic internet that avoids a single platform gatekeeper, while emphasizing taste, presence, and discipline as enduring human advantages.
Get the full analysis with uListen AI
If perceived productivity is high but realized productivity is flat, what changes should leaders make to workflows (smaller diffs, stronger tests, staged rollouts, AI reviewers, stricter definition of done)?
Get the full analysis with uListen AI
Transcript Preview
You've been coding for hours and hours, and he didn't even know what the output was. He was just impressed by the fact that someone could be so locked in. And so I think programming taught me that. It taught me how to focus. It taught me to be disciplined. It taught me to receive this negative feedback from the compiler and overcome it. I do think, you know, we'll need to find what the next version of that is, 'cause I don't think it's gonna be programming necessarily. [upbeat music]
All right, Guillermo, thank you so much for making the time for this. I'm really excited to chat with you today.
Same. People are excited on X as well, so. [chuckles]
So I want to get into codegen, uh, but before we go there, can you talk about what you worked on before Vercel, you know, with Next.js, other projects, and maybe how that fed into what you've built at Vercel?
Yeah. I had a startup before Vercel that I exited to Automattic, the parent company of WordPress.com, and it was quite a successful journey for me because it was my first startup. It's nice to have an exit, but also one of the meta things that I learned... I was a CTO. One of the meta things that I learned is, as a CTO, how can you, you know, influence your team in the best possible way? Like, they are gonna engineer the right things, they're gonna have the best tools. The one thing I did that was revolutionary for my team was spending a lot of time in getting the CI/CD process, meaning continuous integration and continuous deployment of the code that they would write, get it as efficient as possible, meaning you write a little feature, you push code to Git, you get a URL back. I built a real-time system. My background was in writing real-time frameworks, so I was obsessed with real-time streaming of data, and so people would, uh, push to Git, and then I would give them this URL that had a commit ID dot- my company was called LearnBoost. I think it was like learnboostdemo.com. And so imagine that you're almost editing the internet in real time. That was the feeling that I wanted to give my employees, and obviously, I did a lot of other things. I chose technology stacks and whatever, but when I would ask my colleagues, like, "What was the thing that, you know, was most impactful to your- in your time here?" It was that iteration velocity, the, the deployment velocity, the tooling being really neatly configured. I always do the exercise in my head of you show up to a new company. Isn't it nice that Apple kind of figured out operating system and hardware, and I give you a new laptop? Actually, you look- you probably- I don't know about you, but I look forward to having a new laptop.
It's the best.
It's so [chuckles] new and nice.
Install uListen to search the full transcript and get AI-powered insights
Get Full TranscriptGet more from every podcast
AI summaries, searchable transcripts, and fact-checking. Free forever.
Add to Chrome