• dev_null@lemmy.ml
    link
    fedilink
    English
    arrow-up
    0
    ·
    6 months ago

    What would prevent an Android app from having “deep integration with AI”? If the AI is in the cloud then it’s all done through normal web requests, which don’t even require a permission, let alone so special allowance from Google.

    • jg1i@lemmy.world
      link
      fedilink
      English
      arrow-up
      0
      ·
      6 months ago

      Even now they’re already leveraging their OS-level control. The Android Authority guys said in their report, “the Rabbit R1’s launcher app is intended to be preinstalled in the firmware and be granted several privileged, system-level permissions — only some of which we were able to grant”. I don’t work at Rabbit, so I don’t know exactly what modifications they’ve done to their AOSP fork, but they’re doing something.

      If I had to guess, I’d say they’ve messed with the power management of AOSP and probably the process scheduling somehow? I say this because the Rabbit R1 is hands down the fastest way to access an assistant that I’ve used. I have a ChatGPT shortcut on my homescreen of my Pixel 8 phone and the ChatGPT app is constantly killed in the background, so often times I go to access the assistant but I have to wait for an app to load. The R1 is instant.

      And that’s without counting the time it takes to face or fingerprint unlock the phone, then tap an icon.

      No, I would have not paid $200 if Rabbit was an app. I have ChatGPT and Perplexity on my phone, I don’t like the experience compared to the R1. I paid $200 for the end to end Rabbit experience.

      Btw, I get that some people don’t mind unlocking their phone, tapping an icon, waiting for it to load, asking a question, then getting an answer. That’s fine. If you’re happy with that experience, then the Rabbit R1 is not for you.