I think current models have demonstrated an advanced capacity to navigate “language space”. If we assume “software UI space” is a subset of the language space that is used to guide our interactions with software, then it’s fair to assume models will eventually be able to control operating systems and apps as well as the average human. I think the base case on value creation is a function of the productivity gain that results from using natural language instead of a user interface. So how much time do you spend looking at a screen each day and what is your time worth? And then there’s this option that you get: what if models can significantly exceed the capabilities of the average human?
Conservative math: 3B connected people x $0.50/day “value” x 364 days = $546B/yr. You can get 5% a year risk free, so let’s double it for the risk we’re taking. This yields $5T value. Is a $1B investment on someone who is a thought leader in this market an unreasonable bet?
Conservative math: 3B connected people x $0.50/day “value” x 364 days = $546B/yr. You can get 5% a year risk free, so let’s double it for the risk we’re taking. This yields $5T value. Is a $1B investment on someone who is a thought leader in this market an unreasonable bet?