That’s a very insightful write up Ashu. My (much simpler and therefore probably inaccurate analysis) is to view foundational AI as analogous to operating systems. Given that analogy two points of agreement in particular stand out :
1. The foundation model game is brutal. The players have no choice but to play given their commitments till date. Your analysis makes me think that a very plausible scenario is where you have more than one winner just like in operating systems.
2. The application layer is currently underdeveloped and the largest potential source of revenue. While it will also be fragmented, that’s where traditional software economies of scale can be exploited.
One other opportunity worth mentioning - enterprise level AI solutions engineering and architecture.
Thanks for the insights, Ashu
Love it
That’s a very insightful write up Ashu. My (much simpler and therefore probably inaccurate analysis) is to view foundational AI as analogous to operating systems. Given that analogy two points of agreement in particular stand out :
1. The foundation model game is brutal. The players have no choice but to play given their commitments till date. Your analysis makes me think that a very plausible scenario is where you have more than one winner just like in operating systems.
2. The application layer is currently underdeveloped and the largest potential source of revenue. While it will also be fragmented, that’s where traditional software economies of scale can be exploited.
One other opportunity worth mentioning - enterprise level AI solutions engineering and architecture.