Max Hodak: Average Is Not Good Enough

Video summary

Science CEO Max Hodak says Helix and faster iteration drive startup speed, while one retinal implant patient read a 300-page novel.

Science CEO Max Hodak opened with the company’s retinal implant, which stimulates the retina using a laser-equipped camera glasses; after clinical trials, one patient read a 300-page novel. His main lesson: startup speed depends on infrastructure. Science built Helix to connect purchasing, experiments, manufacturing and costs—for example, one wafer iteration cost $40,000—and argued that budgeting and procurement systems prevent delays and expose how quickly research burns cash. Hodak also described a four-stage hiring process, with companywide applicant votes, screens, homework and interviews, plus recurring “IGEN reviews” that aggregate employee feedback. He said fast iteration compounds into an advantage, but founders cannot delegate judgment or rely on universal startup rules. In Q&A, he advised raising twice the cost of the next key experiment, discussed AI’s uses in coding and regulatory work, and said neural engineering could deliver major health and longevity gains. He closed by stressing that biotech is capital-intensive and slow, but can profoundly improve patients’ lives.

Chapters

1. 0:00Infrastructure at Startups: Max Hodak’s Duke Monkey-Neuroscience Experiment
2. 0:51Science’s Retinal Implant: Infrared Stimulation and a Patient’s 300-Page Novel
3. 3:25How Your 17th Employee Buys Things: Budgets, Vendor Delays, and Bulk Materials
4. 7:03What Does an Experiment Actually Cost? Helix Tracks a $40,000 Wafer Iteration
5. 9:31Building a Rigorous Hiring Process: Network Recruiting and Companywide Applicant Votes
6. 11:42Judgment, Horsepower, and Agency: Science’s Four-Stage Hiring Process
7. 16:06Rethinking Performance Reviews: Helix’s Four-to-Six-Week IGEN Reviews
8. 18:53Rate of Iteration Separates Success From Failure: Weekly Learning and Startup Systems
9. 20:38You Can’t Delegate Your Judgment: Founders Must Make the High-Stakes Call
10. 22:47Action Produces Information: Company Systems Can Unblock Growth or Lose Control
11. 25:10Q&A: Formula SAE evidence, thinking-based engineering interviews, and Hodak’s five years at Neuralink
12. 29:36Q&A: Neuralink judgment, human BCI consciousness studies, and more sensitive opsins
13. 35:11Q&A: Helix for AI agents, regulatory automation, and replacing Greenhouse
14. 39:31Q&A: BCI as longevity medicine, rapid stimulation effects, and review-vote checks
15. 42:58Q&A: Raise twice the cost of the next experiment and pursue sustainability
16. 47:46Q&A: BCI packaging, wireless retinal-implant power, and the Stanford-to-Science path
17. 52:15Q&A: Science’s 30/70 culture, six-country trials, and Helix’s vibe-coded purchasing system