Personal agents won't be dominated by a top three
The author argues the future holds multiple coexisting personal assistants, with only a few like Muse and Instinct likely to succeed, and says builders should start from the largest-TAM work use case rather than a generic agent.
In the past 30 days since this post, it is increasingly clear that it will be a world of multiple personal agents and not top 3 ruling the world. The next 30 days may look different but increasingly unlikely. It is still very very hard to build a business around personal agents - we believe personal agent experience should feel like the best outcome to support a large consumer need with a massive TAM, instead of creating a personal agent and applying to every possible use case... a few top ones like Muse and Instinct may succeed at that perhaps. So start with the use case ..the work..the labor ..choose the largest TAM with the most cumbersome work/labor...and the personal agent experience is the outcome of solving it for the users. Don't shy away from human-in-the-loop for advanced needs as that might eventually become your defensibility.
This post quotes another post: x.com/vaibhavbetter/status/2101580076080906724