The 10 Biggest Days in AI History
NLW counts down ten days that shaped generative AI, from OpenAI's founding and ChatGPT to DeepSeek, the agentic era and Anthropic's IPO, arguing each reshaped the industry's structure or power balance.
Watch on YouTube ↗Charlie's summary
- ▶ 2:09
OpenAI was founded out of fear of Google, then needed a for-profit arm and Microsoft's money within four years
NLW says OpenAI launched on December 11th, 2015 because its founders worried about Google reaching AGI first. Elon Musk had already invested about $5 million in DeepMind to keep an eye on it. The backers' initial billion-dollar pledge was “more for the press announcement than anything else.” On June 22nd, 2019, OpenAI took a billion dollars from Microsoft and started a for-profit arm, setting off many fights to come.
- ▶ 3:46
Google invented the transformer but did not see its potential; OpenAI did
The 2017 paper “Attention is all you need” from Google Brain unlocked modern LLMs. Google used transformers for search queries and Google Translate, while OpenAI quickly realized the architecture was what it had been looking for. NLW sees a pattern: other companies had an early advantage over OpenAI and chose not to use it. Noam Shazeer left Google in 2021 after it refused to release a chatbot he had built.
- ▶ 4:49
ChatGPT exists because of a rumor: OpenAI rushed it out to avoid being scooped by Anthropic's chatbot
Anthropic had an internal chatbot, Claude, by February 2022 but sat on it for fear of kickstarting an arms race. In November 2022, a rumor that Anthropic was about to release it pushed OpenAI to ship its own chatbot. Staff expected low tens of thousands of users, and the most optimistic person, Sam Altman, thought it might eventually reach a million. It took 5 days, and 100 million users took 5 weeks.
- ▶ 6:27
The Altman firing reshaped the industry by exposing OpenAI's fragility and pushing Microsoft to hedge
NLW says the November 2023 coup did more than create drama. OpenAI, which had held the best model for more than six months, suddenly looked less unimpeachable, and the returning Altman was never as powerful. Microsoft's contract gave it OpenAI's models only until the board declared AGI, which looked far riskier after the board's behavior. That set in motion Microsoft building its own AI division and slowly moving away from OpenAI.
Give your email to unlock the rest of the key points — for every episode.
Do more with Charlie
Free — sign in with your email, no password.