Most of the planet hadn't heard of OpenAI until like three years ago and now it's a household name.

Most of the planet hadn't heard of OpenAI until like three years ago and now it's a household name. But do you know the whole crazy story behind all of it? Let me tell you in. This is part one in a series explaining the most valuable startups that you can't invest in yet. If you follow startup lore, you know the OpenAI was found in 2015 by Sam Altman and Elon Musk as a nonprofit. For the first five years it was a research lab running experiments. In 2019, they converted to a for profit company. A decision that Musk years later would sue them over. In 2020, they released GPT3. Think of it like autocomplete for anything. You input the beginning of a sentence. It makes sense of it by splicing each word into individual tokens, then using context, lots of it, train on books, websites, Wikipedia, code, almost limitless data points. To guess the end of the sentence, it scores the percentage likelihood for each word and picks the one with the highest score, revolutionary. But the true function point came in 2020 when ChatGTP was released. There are two key differences with ChatGTP. One, it had a really simple consumer UI where you could ask it your question simply. And then two, it was trained by real humans to actually be helpful, not just to predict. For the first time, AI really broke through. 100 million users in 60 days. Open AI then became a fundraiser machine. Today, they raised 122 billion. 50 from Amazon, 40 from Softbank, 30 from Nvidia, 13 from Microsoft. It's the largest private fund raise in tack ever. Why so much cash though? Because building these models is extraordinarily expensive and the cost compounds with every generation. Training GPT3 costs around 5 million. GPT4 cost 100 million and that's just training. Running the models cost roughly 14 billion a year in compute costs, year over year over year. Today, open AI makes money three ways. 45% from consumer subscriptions, 15% from APIs and 40% from large enterprises. This year they're valued at 852 billion and IPO is reportedly on the horizon. But the real finish line isn't going public, it's. In this race they're competing against Google, Meta, and a bunch of others, all spending at the same scale and reaching toward that same finish line. We'll have to follow along to see who wins and what that's going to mean for the rest of us. What tech company on this list should be on pack next?