The Best Engineers in the AI Era

At Sierra Ventures, conversations with engineering leaders help us understand how AI is reshaping software development inside companies building at massive scale.

Anne Gherini recently sat down with Lindsey Simon to discuss what changes when code becomes abundant, why management now requires staying technical, and what separates great engineers from average ones in an AI-native world.

Lindsey has had a front-row seat to multiple platform shifts. He was early at Google App Engine in 2007, worked closely with Brett Taylor at Quip, and joined Vercel in 2020 when the company was still around 15 engineers. Today, Vercel has scaled into one of the most important developer platforms in AI-era software.

His core message was simple: AI changes the mechanics of software engineering, but not the importance of ownership, judgment, and effective thinking.

AI Makes Code Cheap. Judgment Becomes Expensive

One of the clearest shifts Lindsey sees is that generating code is no longer the bottleneck.

“There is a plethora of code coming out,” he explained. “Once there were digital cameras, surely there were more pictures.”

That abundance changes the role of engineering teams. In the past, software engineers spent enormous energy on implementation. Today, AI can generate working code quickly. The challenge becomes evaluating whether the system is solving the right problem in the right way.

According to Lindsey, the best engineers now are the ones who can clearly express intent, define success and failure conditions, and recognize poor design decisions before they compound.

“The better software engineers will know how to express those ends and evaluate those means,” he said.

That shift is also changing how teams think about testing.

Why Test-Driven Development Matters More in the AI Era

Lindsey believes AI has dramatically increased the value of test-driven development.

Historically, many teams treated tests as something written after the code. But AI agents work better when the desired outcome is defined upfront.

“If you write the success and fail criteria first, you’ve really set up a coding agent to succeed and move quickly,” he explained.

At Vercel, the focus is increasingly on building systems that allow agents to continuously validate themselves through fast feedback loops, automated checks, and guardrails.

In practice, that means “vibe codable” systems are not just AI-friendly repositories. They are environments designed for speed, iteration, and rapid verification.

What Makes a Codebase “Vibe Codable”

Lindsey described vibe coding less as a meme and more as a product design principle for engineering systems.

“The premise of vibe coding to me is that you feel like you’re in the flow,” he said.

Anything that slows developers down breaks that flow. Long compile times, slow TypeScript checks, or heavyweight infrastructure steps all create friction for both humans and agents.

The teams that win, in his view, will design environments where AI systems can quickly test, validate, and recover from mistakes automatically.

That philosophy mirrors Vercel’s original product vision around zero-config infrastructure and what Lindsey calls “the pit of success,” where the default path is also the easiest one.

The Best Engineering Managers Still Need to Build

One of the strongest themes from the conversation was Lindsey’s belief that leadership in engineering now requires staying hands-on technically.

“You can’t lead without demonstrating skill in the area that you work in,” he said.

As AI changes programming workflows, Lindsey believes managers can no longer rely solely on past experience or organizational oversight. They need firsthand experience using modern tools themselves.

“I can’t just read about it or talk about vibe coding. I actually have to do it.”

That philosophy has pushed him to spend significantly more time programming again over the last year, despite leading large engineering organizations.

For Lindsey, credibility comes from being “in the trenches” with teams as workflows evolve in real time.

Why Great Engineers Think Before They Code

Despite all the AI hype, Lindsey repeatedly came back to one principle: effective thinking.

He credits much of his leadership philosophy to his interdisciplinary background studying philosophy, English, and physics at the University of Texas.

“The how is more important than the answers,” he said while discussing the concept of “effective thinking.”

That mindset now shapes how he interviews engineers.

Rather than focusing purely on coding skill, Lindsey looks for people who can break problems down clearly, ask strong questions, and reason through tradeoffs before jumping into implementation.

“What questions do you ask? How many things do you figure out before you start coding?”

AI may accelerate implementation, but it amplifies the value of structured thinking.

Why AI Is Raising the Bar for Engineers

Lindsey does not believe AI lowers hiring standards. He believes it raises them.

The engineers who stand out are not simply good coders. They demonstrate ownership, understand business context, and can explain why their decisions mattered.

“I’m not looking for someone who is a good craftsman at writing code,” he said. “I’m looking for somebody who exhibits exceptional ownership.”

That philosophy also changes how Vercel evaluates candidates during interviews.

The company allows AI usage in interviews because Lindsey views AI-native development as the new normal.

“I don’t think software engineering without AI is a viable idea,” he said.

The important signal is not whether someone memorizes syntax. It is whether they can think effectively, collaborate with AI systems, and make sound engineering decisions under ambiguity.

What Lindsey Simon Looks For in Founders and Startups

Outside of Vercel, Lindsey has also been an active angel investor for years.

His framework is surprisingly simple: back people with strong intuition, commitment, and evidence they deeply care about the problems they are solving.

“I look for people that I think are really committed,” he said.

He is especially drawn to developer tools and product engineering infrastructure because those are areas where he has personal intuition as both a builder and user.

Interestingly, many of the strongest signals he looks for mirror what he values in engineers: ownership, curiosity, and demonstrated action over polished narratives.

AI Will Make More People Creative

Lindsey ended the conversation on an optimistic note.

He believes AI dramatically expands who gets to build.

He described showing Vercel’s AI tooling to someone in business development who had never used modern AI creation tools before. Watching them instantly visualize an idea that previously would have required an agency or technical team reinforced how transformative these systems are becoming.

“I love the opening up of that possibility,” he said.

For Lindsey, AI is not just about productivity gains. It is about unlocking creativity and agency for far more people than traditional software development ever allowed.

At Sierra Ventures, conversations like this are part of our ongoing work with engineering and product leaders building AI-native systems inside real production environments. The next generation of software teams will not just write more code faster. They will rethink how software gets designed, reviewed, validated, and built altogether.

Watch the video: