I just saw someone post a question on Reddit about Hoover with regard to the r1. I haven’t heard about this “Hoover” before. Apparently Hoover is a new option that unites all the LAMs, with a subscription that will supposedly cost €50 per month.
Has anyone heard about this? Can you point me to sources?
How does/will this relate to the r1?
There has been discussion on the R1 Discord channel on Hoover, and a few live presentations on Discord. I briefly joined one this morning but had to drop quite soon to leave for work. My understanding so far:
In the presentation that I briefly joined I saw a demonstration where a pong game was generated by hoover.
Direct link to the Jesse’s discord post on this and other developments here.
I’m sure others will correct me if I got (some of) this wrong. 
All will be revealed! It’s all very exciting, but there are still things that aren’t set in stone. We will continue to share sneak peeks here and there as we build things out.
Is it possible Simon that with more sneak peaks and whatnot, rabbit might share those updates with more users? or at least through more avenues than random late night discord calls that are easily missed :-\
@simon
I agree with @MrPickles. I’ve been digging through all of the info on the Discord channel, and it really feels to me that we have a serious information gap in this part of the community. Of course, there is a lot of chit-chat on the Discord channel that I wouldn’t want to see repeated here, but I do believe that we deserve summaries of the news shared about developments. Hoover is a hot topic on the Discord channel, and there has been no mention of it here at all (!?!). We’re not even seeing most of the sneak peaks that have been shared on Discord!
Suggestion: Why not make recordings (or news posts) and share them here, so community members in other time zones (or who don’t have the time to participate in Discord) are kept updated. Don’t we deserve to get the same info?
I think @simon does a great job at making sure that official statements are shared both here and there. Early stage ideas and thoughts get shared there, because it’s a low effort way of getting realtime feedback on things that are too rough around the edges to release.
Personally, I see discord more as “a place where r1 fans hangout, and employees occasionally drop in”
Thinking about the lifecycle of a new feature/idea, I’ve typically seen
1- Someone at the company has a cool idea (or sponsors a community request)
2- We work a quick PoC and show the rest of the team
3- It goes to our alpha testers while we work on it
4- It gets demo’s in discord
5- We bring it up here in the forum to get more feedback
6- We do a press release, because the idea is now a feature
That’s a long list of steps, and ideas get filtered as we make our way through that process - some things get put on hold after the PoC, some things get demo’d but we realize taking it to production isn’t a priority right now, etc.
If you want to see and hear about 100 things we might do, but will likely kill off 75 of them…discord is the place.
If you just want to focus on the 25 that will make it into prod shortly, this is the place 
As with everything in the past of rabbit, I really like to see this without subscription! 50 € sounds pretty insane to me, not sure where that source came from (maybe just community speculating?)
It would be for advanced features.
The source is live session with Jesse on Discord, tossing around ideas related to Hoover.
75% of ideas on Discord don’t make it to an actual release.
I’m also hoping that the r1 will remain without subscriptions.
I would really take pricing with a grain of salt right now, we are far from final decisions.
I will say that with Hoover… it is a very powerful service and one that is not cheap to run for us as a business. How we address that, is obviously a high level business decision that they are currently thinking very hard about.
@simon - I agree and also understand that a Hoover service may require a significant amount of resources to operate.
I have a question and suggestion (at the end of this post)
The concern
The impression that I’m getting is that all of the work currently performed on Hoover is at the expense of other functionality that was promised / announced in this forum. Is this being pushed back (deprioritized)?
E.g.,
The examples I listed are for core r1 functionality, functionality that is keeping existing users onboard and attracts new buyers.
Question / suggestion:
Is it possible to publish a roadmap, so that we can get a better understanding of the committed features that are in the pipeline? This would also help us see features that are prerequisites for later features.
Yup, already took it with a grain of salt, nice to have that doubt confirmed!
Overall I understand that running services is costing money. So far, the selling point of r1 was no subscription.
I know you need to secure steady income and hope you will find a good way.
I agree with the previous post that hopefully, the announced (and more) features for r1 can be slowly, but steadily rolled out! 
I am always looking out for next updates and hope that the current slowdown (after the already happened slowdown for LAM and teach mode development and improvement) will gain pace again.
Please don’t forget about accessibility for blind people, with menu item pronunciation and audio for certain functions.
And by the way: If the promise of staying free of monthly fees is broken so soon, and they introduce the mentioned fee of €50, I see that as a disqualifying factor.
They indicated the subscription would only be for what’s essentially bonus advanced features, and would not impact normal usage or offerings on the r1. As Simon said, it’s a very powerful service, so it’s understandable for something powerful yet optional, there’s a price.
At this point in time, there won’t be any subscriptions for normal use on the r1. No need to start jumping to conclusions based on undecided or early info.
This is why I wish rabbit would be a little more open, even when sharing things that may or may not make it in, it’s like a bad game of telephone unless you catch the discord calls.
It’s a fair concern to have, so let me be as candid as I can and try and answer these questions:
- Is the work on Hoover coming at the expense of other things?
Answer - yes and no.
When it comes to more complex products in particular, sometimes you do have to essentially put all hands on deck to get something out of the door. You saw this last year with LAM Playground and Teach Mode, but after we got those out of the door, we were able to deliver lots of highly requested things, like language support, generative UI, generative voice, and Magic Gallery - all things highly requested by the community.
So I think it’s fair to say that yes, right now, a bit more work is going into Hoover and a bit less work is going into… things at are not Hoover, but that’s just what we need to do right now.
- List of features & functionality question:
Of the things you listed:
Memory Recall improvements are being internally tested and we are seeing a big improvement with them. I don’t currently know what’s needed to get them to production state, but it seems a significant chunk of the work is done.
Touchscreen support and UI 2.0 is all part of the same project. It’s happening, but it’s early days.
Local photo storage has already happened, unless I’m misunderstanding what you mean. Once you enable Magic Gallery, there’s a local copy that syncs to cloud, and you take photos offline and they’ll generate the Magic Camera version later on when there’s a connection again.
- Is it possible to publish a roadmap?
Sure, it’s possible. I’ve managed roadmaps at maybe every startup I’ve ever worked at - I even managed to have somewhat of a public roadmap at Twitter for the primary product I worked on, which was really difficult to get done in a company that big, but I pulled it off! TLDR, I love roadmaps.
But the question isn’t really whether it’s possible, the question is how useful it is. It cannot be prescriptive, because plans change and priorities shift all the time. It has to be vague, it has to be highly qualified, and it has to not create more problems than it solves in terms of user expectation.
And so, for me, as the person that would be in charge of it, if we did it, to put my thumb on the scale right now, it’d be placed somewhere that’s way too vague for me to be comfortable doing it.
There are several reasons for that, one of them being the aforementioned “more resources on less stuff” thing - there are lots of things that both you guys and us as staff want on there and that you could make a case for being on there, but that case is very often not more than “yeah we’ll probably do this but we don’t know when”.
And the other reason is that AI development moves so fast that what isn’t possible one day might be possible the next. That’s what part of what makes it fun and exciting, and in our position, we need to react quickly, and that means an even higher chance of shifting priorities.
When I was at Monzo, our CEO Tom used to say “sometimes, you’ve got to eat your vegetables” - which as a vegan was a very confusing phrase personally since I am always eating vegetables 
- but what he meant was, you will have periods where you spent most of your resources fixing the foundations, making improvements, etc - as opposed to building out new features.
And I believe we will get to that place, which is also a place where it’s easier to have a roadmap. Getting to that place is a good thing, as it typically means you have a fairly solid moat.
But right now, in this market, everyone is in innovation mode, because the market is frothy and fast moving. You don’t want to be ahead of your time, and you don’t want to be late. You want to be exactly on time. And IMO, if we stick the landing with Hoover, that’s where we’ll be and a lot of things will make more sense externally in hindsight
@simon
Thank you for your thorough response and putting this into a larger perspective. I really appreciate that.
That all sounds good. I think, just my personal opinion, that €50 is way too much. I understand that the company needs to make money, and it’s fine that this should come from additional services rather than the basic services on R1. But I’m worried that now things that were promised to us and aren’t even fully developed yet, or not good enough, will only be improved in “pay mode.” For example, LAM Playground could still use a lot of improvement. I always have problems with it, as it doesn’t scroll down properly on the website when online. In its current state, it’s pretty much unusable for most things. Unfortunately. Also, follow-up questions on the target page and other things are awkward. I’ll keep waiting for updates and try to look forward to them.