AI & Machine Learning
Business Insiderabout 2 hours ago
3

DoorDash's CEO says AI coding is not enough to make engineers more productive

AI

DoorDash CEO Tony Xu says AI coding tools only address a fraction of an engineer's work, and true productivity gains require AI integration across all business operations.

DoorDash's CEO says AI coding is not enough to make engineers more productive

Intelligence Insights

Context + impact, normalized for TechCulture.

The Big Picture
In a recent podcast, DoorDash CEO Tony Xu argued that AI-assisted coding, while useful, only covers about 25-50% of an engineer's time spent writing code. He emphasized that engineers also spend significant time on product reviews, design meetings, and cross-team alignment, areas where AI integration is still lacking. Xu noted that DoorDash now generates roughly two-thirds of its code using AI, but this alone hasn't changed the company's organizational structure. He advocates for making the entire company 'AI native' in how it operates, including physical AI applications like food-preparation robots and autonomous delivery vehicles. Xu joins other tech leaders, such as Wix CEO Avishai Abrahami and former OpenAI member Andrej Karpathy, who have publicly highlighted the limitations of current AI tools in delivering transformative productivity gains.
Why It Matters
DoorDash's CEO argues that AI coding tools only address a fraction of an engineer's work, highlighting a critical gap in AI adoption: automating non-coding tasks like design reviews and cross-team alignment. This suggests that companies expecting massive productivity gains from AI must rethink entire workflows, not just code generation, or risk falling short of the hype.

Deepen your understanding

Use our AI to break down complex signals.

Select an AI action to generate more depth.

Tony Xu
Tony Xu
Xu said that coding is only a small part of an engineer's day.

Bloomberg/Getty Images

  • Tony Xu says AI only partly helps DoorDash engineers.
  • Xu said he wants to see AI integration across all departments, not just engineering.
  • AI has boosted DoorDash's productivity, with two-thirds of code now generated by models.

Vibe coding is just one small part of the engineering puzzle, says DoorDash's CEO.

"Only certain parts of a software engineer's day are writing code, and it's great that we have models today to write the code for us," Tony Xu said on an episode of the "Uncapped" podcast released Tuesday. "That helps maybe the whatever 25%, 30%, 50% of our time that's actually shipping code."

Xu cofounded the food and grocery delivery giant in 2013. DoorDash went public in 2020, one of the year's biggest tech IPOs.

The CEO said engineers spend another big chunk of time in product reviews, design meetings, and aligning with different business teams. Those aspects require AI too.

"If that doesn't change and come together, you're not going to be able to just have perhaps the productivity gain that you hope to have," he said. "A bunch of companies across the industry are trying to figure out and get those workflows right so that you're not just AI native from code development, but you're AI native in how you actually operate."

For example, the company still needs to work on parts of its business that can be improved with physical AI, such as food-preparation robots and autonomous delivery vehicles, Xu said.

Xu has taken a moderate approach to talking about how AI has changed his company.

During a May earnings call, he said that AI has not yet changed DoorDash's organizational structure.

"We're seeing a lot of productivity gains right now from AI, about well north of half of our code, probably closer to two-thirds of our code is written by AI today," Xu said. "But that doesn't alone articulate how workflows and team setup ought to change."

He joins a handful of tech leaders who have spoken publicly about the limitations of AI and its usefulness to a company.

Earlier this month, Avishai Abrahami, the CEO of Wix, a web design company that acquired the vibe-coding platform Base44, said the tech industry is overselling AI.

"I would love to say the team in three years, we will be a thousand people in Base44 and a thousand people in Wix, and everything else would be AI agents. I'm not so sure it is possible," Abrahami said. "I think that we all give too much credit for AI and what it can do."

Even the man who coined the term vibe coding has been public about its many limitations for engineers.

In an April talk, former Tesla AI head and OpenAI founding member Andrej Karpathy said that code written by AI is still messy.

"Sometimes I get a little bit of a heart attack because it's not like super amazing code necessarily all the time," he said. "It's very bloaty, and there's a lot of copy-paste, and there's awkward abstractions that are brittle, and it works, but it's just really gross."

Read the original article on Business Insider
Big Tech Developer Tools AI Productivity

Intelligence Exchange

0

Log in to participate in the exchange.

Sign In

Syncing Discussions...

Finding Related Intelligence...