Hacker: AI coding tools perform poorly on large codebases and create unfixable tech debt
Severin Hacker · Duolingo Co-Founder, Severin Hacker: How AI Impacts the Future of Work and Education · 20VC with Harry Stebbings · May 19, 2025 · at 12:55
Severin Hacker is the Co-Founder and CTO of Duolingo. He shares his technical observation after testing AI software engineering tools such as Replit and Lovable.
“On the software engineering side, I think what they're really good at is going from zero to 80% for a simple app. You know, it's like, it's the replet, the lovable, and then the larger the code base, even the one that generates itself, the larger the code base gets, the worse they get. So then, you know, it's just kind of like adding the next 10% Takes as long as the first 80%, and they kind of create their own tech debt which they also cannot really solve.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →