Benchmark GP, Victor Lazarte: The 3 Traits All the Best Founders Have
Watch on YouTube ↗ Summary based on the YouTube transcript and episode description.
Benchmark GP Victor Lazarte argues AI will destroy 99% of knowledge work and that replacing humans is the most exciting venture opportunity right now.
- Lazarte predicts ~99% of knowledge work as defined today will not exist in 10 years, with founders and shareholders capturing most of the wealth.
- He invested in Brex pre-product after helping founder Pedro Franceschi escape a predatory Brazilian investor deal; Brex hit $100M revenue in 18 months.
- Mercor grew from $1M to $100M+ ARR in 11 months; Lazarte invested when they had $1M revenue because AI interviewers will outperform humans on a hard, defensible problem.
- Key founder filter: open-minded but highly disagreeable simultaneously — genuinely curious yet willing to upset you with their contrary conclusion.
- LLM inference compute grew ~100x in two years; Lazarte uses “does your company get better or worse as models improve” as his primary AI revenue quality test.
- Benchmark’s first check from Lazarte was $55M into HeyGen from a $600M fund — fund size is explicitly not a portfolio construction constraint.
- Lazarte recommends studying computer science now for the same reason you’d study math after calculators: transfer learning and structural thinking remain valuable.
- Duolingo is his 10-year public stock pick — he believes founder Luis von Ahn will build a free AI tutor that becomes one of the most valuable companies in the world.
2025-04-14 · Watch on YouTube