Jacob Lauritzen, speaking on the 20VC podcast this week, made a blunt observation: AI has flipped the engineering productivity equation. The bottleneck is no longer writing code—it's reviewing it. The conversation, covered by Crypto Briefing, offers a window into how software teams are reorganizing around a new reality.
The bottleneck moves
For years, the slow part of shipping software was the typing. Engineers sat at keyboards, cranking out lines. Lauritzen says AI tools have changed that. Now code generation is fast—almost too fast. The drag becomes what happens after: getting that code reviewed, approved, and merged. Teams that don't adapt to this shift risk piling up pull requests and burning out reviewers.
From creation to design
Lauritzen's second point ties directly to the first. If AI handles more of the grunt work, engineers need to spend their time on higher-level thinking. Systems design—architecture, trade-offs, how pieces fit together—becomes the core job. The implication is blunt: engineers who only write code will matter less. Those who can design systems, and review others' work critically, will be the ones who drive velocity.
Crypto Briefing's coverage of Lauritzen's talk lands in a sector that runs on tight engineering cycles. Smart contracts, exchanges, DeFi protocols—they all ship code that has to be near-perfect. If the bottleneck is now review, crypto teams face a particular pressure: the reviewer needs to catch not just logic bugs but economic or security flaws. Lauritzen didn't say it outright, but the implication for any crypto project is clear—invest in your review pipeline or get left behind.
The interview leaves an open question: how fast can engineering orgs retool their workflows? For now, Lauritzen's core message is already being tested by teams that ship daily.
