Returns for our first round of investors are capped at 100x their investment (commensurate with the risks in front of us), and we expect this multiple to be lower for future rounds as we make further progress.
I remember OpenAI or Sam Altman saying that in subsequent funding rounds, the profit cap would decrease, eventually reaching 2x. But I can’t find a source for this right now. (Even if I am remembering correctly, who knows if OpenAI ever actually lowered the cap below 100x.)
Now that I’m thinking about it more, the 100x profit cap was always too low. If OpenAI’s valuation in 2019 was $3 billion or less, then now, with a valuation of $300 billion, those first-round investors would already hit the cap on their returns.
It seems like OpenAI was already trying to rectify this problem before its recent announcement. In 2023, The Economist reported that:
Profits for investors in this venture were capped at 100 times their investment (though thanks to a rule change this cap will rise by 20% a year starting in 2025). Any profits above the cap flow to the parent non-profit.
If the purpose of the profit cap (or return on investment cap) is to limit OpenAI investors from having an obscene level of ownership over the wealth generated by AGI, this this makes a lot more sense. (I’m assuming that the profit cap is abandoned now and this is a moot point, but I find it interesting to think about anyway.)
If OpenAI had stuck to the plan of increasing the 100x profit cap by 20% year every year starting in 2025, here’s what the profit cap would be in future years.[1]
2030: 250x
2035: 620x
2040: 1,540x
2045: 3,830x
2050: 9,540x
2055: 23,740x
2060: 59,070x
2065: 146,980x
2070: 365,730x
As time goes on, eventually the number gets too big, but even 365,730x is not a totally unprecedented return on investment in the pre-AGI world. Mike Markkula’s angel investment in Apple would have had a return on investment (ROI) of over 3,000,000x had he retained his shares from the beginning until the 2020s.
If you look up lists of the stocks that have grown the most from IPO to their all-time high, or the venture capital investments that have had the best ROI, you see some numbers in the 1,000x to 10,000x range. So, a 10,000x cap would not be unreasonable.
If you think the amount of wealth generated by AGI will be essentially unlimited and defy calculation by conventional standards, then it shouldn’t be a problem to have a cap of 10,000x or even 100,000x or 1,000,000x, since that would still end up being a small percentage of the overall amount of wealth generated by AGI.
From the 2019 announcement:
I remember OpenAI or Sam Altman saying that in subsequent funding rounds, the profit cap would decrease, eventually reaching 2x. But I can’t find a source for this right now. (Even if I am remembering correctly, who knows if OpenAI ever actually lowered the cap below 100x.)
Now that I’m thinking about it more, the 100x profit cap was always too low. If OpenAI’s valuation in 2019 was $3 billion or less, then now, with a valuation of $300 billion, those first-round investors would already hit the cap on their returns.
It seems like OpenAI was already trying to rectify this problem before its recent announcement. In 2023, The Economist reported that:
If the purpose of the profit cap (or return on investment cap) is to limit OpenAI investors from having an obscene level of ownership over the wealth generated by AGI, this this makes a lot more sense. (I’m assuming that the profit cap is abandoned now and this is a moot point, but I find it interesting to think about anyway.)
If OpenAI had stuck to the plan of increasing the 100x profit cap by 20% year every year starting in 2025, here’s what the profit cap would be in future years.[1]
2030: 250x
2035: 620x
2040: 1,540x
2045: 3,830x
2050: 9,540x
2055: 23,740x
2060: 59,070x
2065: 146,980x
2070: 365,730x
As time goes on, eventually the number gets too big, but even 365,730x is not a totally unprecedented return on investment in the pre-AGI world. Mike Markkula’s angel investment in Apple would have had a return on investment (ROI) of over 3,000,000x had he retained his shares from the beginning until the 2020s.
If you look up lists of the stocks that have grown the most from IPO to their all-time high, or the venture capital investments that have had the best ROI, you see some numbers in the 1,000x to 10,000x range. So, a 10,000x cap would not be unreasonable.
If you think the amount of wealth generated by AGI will be essentially unlimited and defy calculation by conventional standards, then it shouldn’t be a problem to have a cap of 10,000x or even 100,000x or 1,000,000x, since that would still end up being a small percentage of the overall amount of wealth generated by AGI.
I used this compound interest calculator to figure this out: https://​​www.investor.gov/​​financial-tools-calculators/​​calculators/​​compound-interest-calculator I rounded these numbers to the nearest ten.