DarkPlayer

Re the whole Rabbit thing: One of my past jobs was for a device based on AOSP (Android Open Source Project). We built a high-level platform, including features like a launcher and push notifications, without needing to write “an OS”. AOSP is an excellent foundation for this, and in the end your customers want you to build the features that make the product useful, not rewrite the basics someone else has already done for you - which would just make the product more expensive. What makes the Rabbit thing different is that it appears to be a rushed product entering into an unproven ecosystem of portable AI assistant devices. There was some expectation that the hardware played a key part of what made it worth all the extra cost and inconvenience to consumers, that it couldn’t have been done without something that’s unique about it. But the app being completely usable on a rooted Android 13 phone made it clear that’s not exactly true. (Root is required to give it some system-level permissions, e.g. to adjust brightness or connect to Wi-Fi.) In response, they issued a statement saying the product is “not an Android app” and that “there are some unofficial rabbit OS app/website emulators out there”. Which is an interesting oversimplification, because next they say that “rabbit OS and LAM run on the cloud with very bespoke AOSP and lower level firmware modifications, therefore a local bootleg APK without the proper OS and Cloud endpoints won't be able to access our service”. Clearly it was working after granting those system permissions on a Pixel 6a. Nothing seems to indicate they need further permissions than a regular app from the Play Store could request, and that’s the problem. They can’t answer why this couldn’t have been an app. They “fixed” this PR issue by issuing an update that now sends the device’s IMEI with every request, which is checked to ensure it’s within the range of IMEIs on sold devices. Only a temporary solution until someone figures out the ranges, but still shows the lack of unique hardware makes it easy to skip buying the product entirely. Does this really actually matter? Well, there’s this whole other fiasco that feels too similar. Juicero was an overdesigned juice maker appliance released back in 2016, after the company developed it for 3 years. It required Wi-Fi, and you needed to buy their juice packs, which had QR codes that prevented 3rd-parties from offering their own (they promised this was just to stop you from using expired juice). The company basically crumbled the moment an article came out saying, you… you can just squeeze the bags yourself and not buy the $399 (initially $799!) appliance. 7 months after the article, the company shut down. Rabbit and Humane give me similar vibes. Juicero was overdesigned and far too expensive for what it was. They refused to believe their product should be simplified and cost-reduced, the price only came down because sales were so poor. They decided to throw away all that work, instead of seeing their product for what it was. As neat as these AI products are in design, I feel they’re far too rushed to market, too unproven as something the average person would find useful, and too much money went into them to sell cheap enough to get meaningful sales. Maybe they’ll become super useful in a few years, or maybe the segment will never take off. Either way, it seems these are niche products targeting a market that isn’t particularly interested in bringing an orange single-purpose don’t-call-it-a-phone with them everywhere, or wearing a camera on their chest. We’ve also all seen how it goes when AI gets stuff wrong and you need to fight it to give you the right output, and that’s just not a good experience. Phones are just all around easier to use for now. I guess we’ll see, and I’d love to be wrong because I hate to see all that work (and the devices themselves) go to waste, but so far it seems these products have a limited shelf life (enforced by a QR code on the packet 🤪) and won’t last that much longer. Also, don’t buy a product that’s clearly unfinished, even when the company insists it’ll be finished with updates. Buy products for what they are right now, and consider any updates a lucky bonus.