Court Exhibit·21 JUL 2017
Plan the Exit Before You Need It: Staging a Mission
The most important strategic decision isn't where you start — it's knowing which structure you'll need next, and when to trigger the shift.
Source document — Musk v. Altman (OpenAI) — Exhibit 16 · Musk v. Altman (OpenAI) · 4:24-cv-04722 (CAND), Doc. 351-19, filed 2025-11-07 (page 1)
Excerpt · In Elon Musk & Greg Brockman & Ilya Sutskever & Jared Birchall's own words
Greg Brockman: "100% agreed. We think the path must be: 1. AI research non-profit (through end of 2017) 2. AI research + hardware for-profit (starting 2018) 3. Government project (when: ??) Government project is helpful for financial resources, but crucial for security — don't know how to defend against Putin or North Korea otherwise. When ready, we must proactively form the project around us, rather than being slurped in against our will."
Elon Musk (responding to a NYT article on China's 2030 AI ambitions): "They will do whatever it takes to obtain what we develop. Maybe another reason to change course."
Elon Musk (separately, to the founding team): "By the way, in appreciation for what you've done to get OpenAI to where it is today, I would like to give you each a Founder Series Model 3. These are the earliest cars produced and are not available to the public. If there are a few others at OpenAI who you think also really deserve one (maybe Wojciech?), I am happy to do the same for them."
Ilya Sutskever: "Wow, thank you! This is a very kind gesture that we greatly appreciate. Yes, Wojciech definitely deserves it as much as we do, as does John Schulman."
Elon Musk: "Consider it done! Jared, please facilitate. Let's try to slot these four cars into next month's production."
Jared Birchall: "Will do. Guys — I'll reach out separately with details."
The daily letter · Free
This edition is for members — but the daily letter is free.
One real corporate email, and the lesson it teaches, in your inbox every weekday morning. Free. Unsubscribe in one click.
How this surfaced
- Source type
- Court Exhibit
- Case / record
- Musk v. Altman (OpenAI)
- Citation
- 4:24-cv-04722 (CAND), Doc. 351-19, filed 2025-11-07
- Date authored
- July 21, 2017
- License
- Public domain
- Original
- View the primary source →
More from Elon Musk
Elon Musk & Ilya Sutskever ·OpenAI ·18 FEB 2016
When A Rival Leaps Ahead, Change Your Cadence
A single alarming data point should reset how often your leadership team talks, not just what they talk about.
Ilya Sutskever ·OpenAI ·12 JUL 2017
Turn A Research Problem Into A Capital Problem
When you can't invent the answer, buy the conditions under which the answer becomes findable.
Sam Altman ·OpenAI ·6 MAR 2019
Design The Structure Before You Need The Money
When your mission and your capital needs collide, invent the vehicle first — then go raise.
Shanna Velez & Mike Marchak & Purnima Kochikar & Alyssa Ablao & Josh Kim & Randy Gelber ·Google ·15 DEC 2021
Discounts Are Cheaper Than Exceptions to Your Pricing Rule
When you can't grant a carveout, invent a program — same economics, but the rule stays intact.
Don ·Google ·14 MAR 2023
Bundle Deals Across Divisions to Save a Weak Position
When one business unit can't justify the check alone, stack the strategic value across the whole company — or lose the partner entirely.
Karen Aviram Beatty & Lawrence Koh ·Google ·11 DEC 2019
Bundling Deals Can Backfire When Your Counterparty Wants Simplicity
When you stack asks to justify a payment, the other side sees complexity — and offers you a simpler exit.