“I don’t recommend what we did for most companies” | Unconfuse Me with Bill Gates

Summary

Sam Altman and Satya Nadella have a wide-ranging conversation about OpenAI's origin story, unique culture, and key lessons learned. They discuss taking an unconventional path in AI research to pursue general intelligence, assembling a talented team, gaining Microsoft's crucial long-term support, the excitement around achieving AGI, and the advice they commonly give others. Altman emphasizes the importance of calibrated risk-taking and pursuing meaningful work, while Nadella focuses on building teams with diverse skills.

Takeaways

• 👥 Having the right mix of skills on a team is critical for success
• 🚀 Taking risks to pursue what you really want can pay off enormously
• 🗣 Communicating clearly what you want can take you far
• 📈 OpenAI started with no revenue plan but had faith the AI would eventually be valuable
• 💡 OpenAI did many things counter to typical startup advice, but it worked for them
• 💰Microsoft's patient capital was key to supporting OpenAI's long-term vision
• ⚡ OpenAI moves incredibly fast given the depth of the technology
• 🌟 Great people want to work with other great people on meaningful projects
• 🥅 OpenAI has stayed focused on its mission to create AGI
• 💎 Exceptional talent plus alignment on a goal enables magical outcomes

Q & A

What was unique about OpenAI's founding and path to market?

OpenAI took an unusually long time (4.5 years) to launch its first product, started with no revenue plan, and did not talk to users early on - very counter to typical startup advice.

Why was Microsoft's backing so important?

Microsoft provided crucial long-term "patient capital" to fund OpenAI's ambitious research agenda pre-revenue.

How does OpenAI continue to attract top talent?

Great researchers want to work alongside other great colleagues on meaningful, cutting-edge projects like AGI.

What drives OpenAI's speed of execution?

The incredible talent working on hard technical problems, while staying aligned to the overarching mission.

What advice does the OpenAI CEO often give?

Pursue what you really want even if it seems risky; you'll likely regret it more if you don't try. And be very clear about what you want.

How did the OpenAI CEO's experience at Y Combinator inform decisions at OpenAI?

Seeing typical startup playbooks and rules made the CEO understand when, why and how OpenAI could intentionally break them.

What surprised the OpenAI CEO about building teams?

The importance of having diversity of skills, not just pure engineering IQ.

How big is the OpenAI team compared to tech giants?

With around 500 people, OpenAI is still a very small team versus companies like Google and Microsoft.

What areas could OpenAI improve on?

Scaling all parts of the organization to match its global prominence, including PR, customer relations, etc.

What makes the OpenAI mission so compelling?

The chance to be part of inventing artificial general intelligence draws top researchers eager to make history.