A Fish Needs a Bicycle

My mother is one of those librarians who wants to teach you something; if you ask her a question she'll usually say "I won't tell you the answer, but I'll show you how to find out!" which is both a great life lesson and sometimes a little annoying.

But it's a lesson we really have internalized at Rover. We're a knowledge layer, to help you make better decisions, to help you think better. We believe that AI is a fantastic tool for helping—helping you remember, helping you understand, helping you see things you didn't before—not a tool for doing things for you. We at Rover are in fully in the 'bicycles for the mind' camp.

So not to throw shade at Muse (OK, a little bit to throw shade at Muse) or any of the others, but to us it's fundamentally not good to have an agent handle your entire life on your (or, really, Zuckerberg's) behalf. The obstacle really is the way, and if you just put your agent to work tackling the obstacle, well, you've certainly offloaded something but you've lost something important in return.

One of my favorite experiences on Rover is when you ask about what someone else thinks about something. For instance, yesterday, my partner Josh asked Rover "What does Mark think about the midterm elections in Kansas?" And Rover was able to send back a nice three-paragraph rundown (essentially, I think the GOP will win the governor's race and all but one House race, but Democrats have a real shot at the Senate; Hamilton is a good candidate and Marshall is....not) that was an accurate reflection of my opinion.

We more typically use Rover to level set on decisions around the company, like "What did we decide about which LLM to use for this function," and Rover is super useful for that, but to me the things that really are magic are the personal ones, the ones that tell you what a person values and believes, not just what he says. Josh's question sparked a nice conversation between us that probably wouldn't have happened otherwise, and we learned something about our selves beyond work. Rover was the lever that multiplied this force, augmenting, not replacing, and for use that's exactly where we want to be.