Rabbit R1's LAM has a moat problem

The recent Google I/O 2024 keynote highlighted some of Google's new integrations of Gemini into Android. While I'm an Apple fanboy myself, the demos showcasing the cohesive integration of Gemini into the native interface were very impressive. Google has set a high bar for my expectations from WWDC, but I couldn't stop thinking about the effects of these integrations in some of the fancy new AI gizmo startups.

The Rabbit R1 is a $199 USD device that features a "Large Action Model" (LAM). Rabbit's LAM hopes to act like a human to interact with external services. However, there have been many concerns regarding their concerning security practices of storing your credentials for other services.

Compared to tech giants like Google, Rabbit lacks infrastructure, data sources, and a huge user base to fuel development and platform integration. Google has a massive edge in this space because they own:

• Android, your phone's OS
• The world's most popular search engine
• Massive compute data centers to train better models
• Gmail & YouTube for personalized experiences

With so many services around, discoverability is king, and companies crave to be the top Google search result. This trend will likely follow into Gemini, where companies will prioritize integrating with Google's platform to be accessible. Unlike Google, Rabbit lacks this ability to leverage other brands into integration with their LAM.

It's clear that tight-knit AI integration into our smartphones will leave Rabbit fighting for scraps among giants. Rabbit has a moat problem, and it needs to leap high and far to win the carrot from users.