Deep learning works and improves with scale; this belief was considered heretical at the time.

I initially aimed to start a robotics company, believing it would be the first successful business built on deep learning, but I ended up at OpenAI to learn from smart people and gain expertise in deep learning.

Deciding to start OpenAI was a pivotal moment for us. We almost didn't do it because AGI seemed like a crazy idea, and there were many other successful startups to pursue. It took a core group of people to look each other in the eye and commit to the mission, despite the doubts.

The Dota 2 project at OpenAI reinforced the belief that scaling experience is key to improving AI capabilities.

The scaling laws became a foundational principle for AI development after being popularized by researchers, including those at OpenAI and Google DeepMind, who emphasized that it's not just about making models larger, but also about training them on sufficient data.

We realized that scaling laws in AI showed predictable improvements with increased data and compute, which encouraged us to focus on building a product for knowledge work as AI matured.

After two decades in AI, I've witnessed the transition from AI winters to the rise of deep learning, which has opened up new possibilities in various applications.

The scaling laws we discovered were crucial to our success. Reflecting on the early days of AI, it seemed improbable that we would achieve what we have now, especially when the idea of chatbots felt like science fiction.

Investing in OpenAI in 2018 felt simple because I had a strong belief in AI, despite the investment being unconventional and lacking a clear business plan.