OpenAI CEO explains how they broke startup rules for ChatGPT
OpenAI CEO Sam Altman recently discussed his company's success during a fireside chat hosted by fintech company Stripe. OpenAI, the artificial intelligence venture behind ChatGPT and GPT-4, is now valued at nearly $30 billion. However, Altman revealed that OpenAI broke all the advice given by the startup accelerator Y Combinator:
- It took four and half years to launch a product.
- OpenAI is now the most capital-intensive startup in Silicon Valley.
- They built a technology without knowing who their customers were going to be or what they would use it for.
Altman tweeted that ChatGPT, a product of OpenAI, had no social features or built-in sharing, required sign up before use, and had no inherent viral loop. He also mentioned that he was seriously questioning the years of advice he gave to startups. However, because of his experience evaluating startups as the former CEO of Y Combinator and CEO of Reddit, he was confident in breaking the rules.
According to Altman, potential investors did tell him that he was doing it wrong, but he didn't care and asked them not to invest. OpenAI's President and Co-founder Greg Brockman reflected on the company's rule-breaking ways and said that they spent a couple of months writing down different ideas for GPT-3 and GPT-4. Instead of following the standard advice, OpenAI decided to ignore it completely and became successful in doing so.
Altman and Brockman agreed that artificial intelligence is just different, and every company, individual, or business is a language business, and it has language flows baked in. If adding a little bit of value in existing language workflows, it will be adopted so broadly.
