← Home
The Twenty Minute VC · November 2, 2026 · 1h 2m

20VC: Who Wins the Inference Market | Has Open-Source LLM Inference Peaked? | What Happens to the Cost of Compute | The Real Limits of Post-Training and RL with Erik Bernhardsson, Modal CEO

Erik Bernhardsson, co-founder and CEO of Modal, joins to discuss the economics and competitive dynamics of AI infrastructure. Modal has raised over $1.2 billion—including a $750 million Series D led by Accel at a $15.75 billion valuation—and surpassed $1 billion in ARR. The conversation covers who wins the inference market, GPU pricing and data-center buildout, custom silicon and CUDA, open-source versus frontier models, and the limits of post-training and reinforcement learning. Erik also reflects on raising capital without a pitch deck, founding in his mid-30s, and where value may accrue for enterprises and startups across the AI stack.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Preview

•
Founding in Your Mid-30s00:05:30
Erik discusses waiting until his mid-30s to start Modal and whether founding a company is really like "eating glass."
•
Raising $1B Without a Pitch Deck00:08:00
Erik covers raising a large round without a pitch deck and the risks of raising too much venture capital.

3 more ideas & all timestamps

This episode is in its early-access window. The full breakdown unlocks free in about 70 hours — Pro members read everything the moment it lands.

Read it now with Pro$10/mo · founding $96/yr
Was this useful?