What's interesting is that almost all of the third-party integrations the company has deployed—DoorDash, Uber, Midjourney, etc.—are down, so those functions don't work. (Not that they actually worked before.) The scroll wheel is less jerky and the interface improvements are 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 the R1 over the past year, but three stand out: Beta Rabbit, LAM Playground, and Teach Mode.
Beta Rabbit uses advanced large language models to deliver a more conversational experience when you ask R1 anything. I don't find it as conversational GPT-4o or Living Gemini. I asked him how we knew 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 that could be detected. R1 then begins a series of “Search for the cosmic microwave background,” “Search for this,” and “Search for that.” After five times like this, it finally started reading out the answer about CMB.
LAM Playground is an exciting feature accessible on Rabbit's web portal. These “large action figures” run on a virtual browser that you can interact with, and are largely intended to showcase how Rabbit can perform tasks on your behalf. (You know, because it can't do that on launch!) Enter the 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, although you'll need to log into Amazon through this virtual browser, which seems like a big risk. about privacy.
I asked it to Google “best office chair” and then took me to the retailer's website. It took so long to do this and it said “best office chair review 2023”. (Is R1 in another timeline?) But it still goes to the first result (which is my own result Instructions for using office chairsthank you very much) and it takes me to the product page Branch Ergonomic Pro ChairMy top recommendation.
Where LAM Playground lets you see how all this works, Teach Mode lets you put it into practice. It's still in beta (I believe the R1 is also still in beta). After a few tries, I kept getting errors that made the feature unusable. It finally worked the other day. I created a lesson and then walked through the steps in a virtual browser—R1 recorded each mouse click. Then when I ask my R1 device to perform an action, it mimics my action.