The funny thing is that almost all the third-party integrations launched by the company – DoorDash, Uber, MidJourney and others – are being discontinued, so those functions don't work. (Not that they really worked before.) The scroll wheel is less janky, and the interface improvement is welcome. For example, you can now press and hold the push-to-talk button and scroll up or down to change the volume.
There have been a lot of updates to R1 over the past year, but there are three major ones: Beta Rabbit, LAM Playground, and Teach Mode.
Beta Rabbit uses advanced large language models for a more conversational experience when you ask R1 anything. It didn't feel like a conversation to me GPT-4o Or Gemini Live. I asked him how we know anything about the early years of the universe, and he began reading an excerpt. At one point, it mentioned the “cosmic microwave background”, at which point I interrupted it and asked how it is detected. R1 then started shouting “discovered the cosmic microwave background,” “discovered this,” and “discovered that.” After five of these, I finally started reading answers about the CMB.
LAM is an interesting feature available on Playground Rabbit's web portal. These “large action models” run on a virtual browser that you can interact with, and their main purpose is to show how Rabbit would be able to execute tasks on your behalf. (You know, because it couldn't do it at launch!) Enter a prompt and Rabbit's bots will execute it. For example, you can ask it to find an item and add it to your Amazon cart, though you'll have to log into Amazon through this virtual browser, which seems like a big privacy risk.
I told it to search “best office chair” on Google and then take me to the retailer's webpage. Took a very long time to do this and typed it into “Best Office Chair Reviews 2023”. (Is R1 in a different timeline?) But it still went to the first result (which is my own office chair guideThank you very much), and it took me to the product page Arm's Ergonomic Chair ProMy top recommendation.
While LAM Playground lets you see how it all works, Teach Mode lets you put it into action. It's still in beta (I would argue that R1 is also still in beta). After a few attempts, I kept getting an error that made the feature unusable. Ultimately, it worked on a different day. I created a text, and then executed the steps within a virtual browser—R1 logs every click. Then, when I asked my R1 device to take action, it mimicked my actions.