If it looked more like R2-D2 or C-3PO, it would sell like hotcakes.
Or Chewbacca, even.
2 Likes
Like a couple of other Librem products, maybe Mimi will be in a future movie 
5 Likes
Good post. I think that was what was missing from the original post i.e. explaining what such a device could do for the customer i.e. making it more concrete.
4 Likes
Translation: You arenāt buying anything. You are funding R&D.
You are supporting the idea of bringing an open alternative into a world that will otherwise be made and controlled by blackbox, surveillance-capitalist corporations ā¦ by funding the development of a product.
I think we can see that Purism has learned from the Librem 5 experience. Effectively Purism has decoupled crowdfunding of the development of a product from actually purchasing the product. (But if you do decide to purchase the product then your funding of the development of the product gives you credit towards that purchase, via the coupon system.)
Purism has also already stated that this crowdfunding campaign will assess the level of interest in such a product (where genuine interest has to be shown by being prepared to put down at least a little of your hard-earned).
Yes, I think we all understand that crowdfunding is inherently higher risk for the customer than just buying an already-existing product from the store. A link above does go over some of the risks (see heading āRisksā). I think it is fair to say, comparing Librem 5 with MiMi, that less of the risk is on Purism now - but it is also the case that the supporter can control his or her financial risk to a greater extent now, again comparing Librem 5 with MiMi.
The supporter can just chuck in $5 or $20 as a one-time indication of support, and leave it at that. So it doesnāt need to be a major financial commitment if you support the idea but donāt have a lot of spare cash and/or would prefer to defer and decouple a purchasing decision entirely until when/if a product exists.
It could easily be that both parts of that are wrong - but you are right to raise the (implied) question.
The more adventurous might be able to achieve that i.e. remove brain and insert into different body. But the product has to exist first.
4 Likes
One question that could be covered by āO. Charging MiMi Batteryā is:
will MiMi be able to plug herself in for charging purposes when the charge level gets low?
Also note section āH. Sensors:ā has typo in ādetectioonā (2 occurrences) @JCS
1 Like
Now that is something completely different from what the original page tells about this. HCA is very different from a robot. Very different designs, goals, systems, problems to solve and usecases. Makes me worried for the project. If thatās the goal, the core idea, then - remembering what I said about the walkman - it would be more sensible to make that sans the mechanized body. The cute hardware is a distraction from the core use idea and adds a lot of cost and complexity with very little in return (for company, in usability, in value for user etc.), as far as I can tell.
What would be a more realistically achievable and financially sensible approach? How a bout a a simple āsoupcan speakerā type ābrainā first (or do other design choise, like a wallframe using an already existing Purism tablet or that hybrid vacuumdroid) and then doing a network of various add-ons (sensors, displays, speaker-mics, droids etc.) that the user can choose to increase assistant system. When thereās a working home assistant system and users, a humanoid droid might come into play.
So, I guess my question is: Why go for the biggest and most complex single monolithic do-it-all system, both as a design choice and as a strategic business decision? [edit: Expanding on the latter, what does a robot do for Purism other than differentiate it, what products come after it and where does it lead - more toys or a more private robot dog (with flamethrower) - and how do those align with the current portfolio. Iām likely not a customer but this affects how much life I see in other products.]
3 Likes
@JR-Fi, What I meant in my post is that the MiMi will be a home computer assistant but not only. In its latest evolution, it will be able to let you command or control it to move around and do things that a static device cannot do. I gave some example applications of what a simple home computer cannot do like remotely entering it to use its eyes and move around in your house to check that everything is OK when youāre not at home. It could do so much more like feeding your fish or your cat every day, watching your baby during their sleepā¦ so much we can imagine a robotic device doing. With so many hardware/physical capabilities, imagination is the limit really.
Regarding you concerns, if you re-check the page you will see that it is a multi-year project with milestones and one can decide to stop supporting when they are happy with the product.
- The first milestone will be a simple board, with sensors and software (mainly for devs)
- The second milestone will be a simple home computer with a screen, cameras, speakers, micā¦ (thatās the head of the robot).
- The third milestone will be the head, with arms, running on wheels.
- The fourth and final milestone will be the humanoid robot with legs.
6 Likes
While true that it probably could do those things (if it can open bags, raise to tables and cabinet shelves etc.), I maintain that the mobile robot is not the most efficient, economical etc. choice to do that. Static sensors maintain constant vigilance in comparison to roaming guard, automatic fishfeeders are already a thing and quite simple (pretty sure thereās one for cats too), nannycams are less intrusive and quiet than a moving droid etc. A robot makes sense if something needs to be moved (and in this case that canāt be very big due to itās size and reach) and it canāt be pre-automated, and for situations that occur randomly and seldom.
The latter points do give a bit of hope that the first milestones could by themselves be useful in the context of smart living. That however needs a largish ecosystem of various compatible devices and a general open control protocol. Developing those would be big. I hope this will be compatible, as that would increase its value (to be able to have a trusted control of other smart devices and have them be able to interact). Robots are cool and have fun with them if you want but they are not what Iād give money for, they are not a solution to actual problem or need.
[edit to add: Make it vacuum as well and give it a flamethrower and I might change my mind. Stretchgoals? 
]
3 Likes
Perhaps there is a generic question here, summarised as: jack of all trades, master of none. A general purpose computer will very rarely be the best at doing any given specialised task, but one general purpose computer can cover a wide range of tasks. Could that logic also apply to a robot of this type? Only the future will answer that.
In other words, sure, as a hypothetical, I can buy a number of cat feeders and a large number of fixed IP cameras (albeit with PTZ) and provide cat feeding services and remote visual surveillance services without even having a computer, much less a robot. Or one robot can in theory do both tasks.
Where a robot would shine is in handling unforeseen circumstances. For example, I have my doubts that there are cat feeders that will order more cat food and replenish the cat feeder if the human carer is unexpectedly delayed by days more than planned (but I stand to be corrected if someone links to such a product).
I donāt doubt that there are some challenges here with these tasks - such as whether a noisy, moving robot would so freak out the cats that they wouldnāt eat anyway - or that as the robot moved clumsily, opening a door, going through it, closing a door ā¦ a cat would shoot through the open door, because we all know that open doors trigger superluminal motion in cats.
I agree with you that this is a big project and I hope that Purism has a good handle on cost and time!
2 Likes
It sounds like milestone 2 would essentially be an Alexa competitor similar to what Mycroft were trying to do before they went bust.
Be interesting to see if any of their work can be repurposed.
3 Likes
If Purism came out with a freedom respecting security camera like the PineCube, or Raspberry Pi + Camera + PoE Hat, except also with
- outdoor & vandal proof rated housing (ideally looks good indoors & outdoors)
- colour/infrared night vision
- decent quality camera sensor + lens
- hardware kill switch for camera & for microphone
- speaker & light
- PoE+ & USB-C power options
- microSD slot for independent operation
- PureOS libre software, perhaps with a package that you install to put it into security camera mode with code from ZoneMinder or MotionEye
- bonus points if the camera has an attachment to let me feed a pet and power via outlet like this robot can do, or if there is a easy way to use it as a baby monitor streaming to a Librem 5/11 for example
- additional bonus points if it could do vehicle, person, pet, package face detection/recognition, and various noise detection like pets, babies, fire alarm, locally on-camera, or on the NVR
Maybe combined that with a 3.5" HDD 64DD-style to the Librem Mini perhaps, using a WD Purple or Seagate Skyhawk for recording, E2EE to the device, and Iād buy a 4-pack. MSRP $1.5k-$3.0k.
Maybe itās just me but I feel a camera baby monitor is a little less creepy than having a robot standing in the room watching. You can also mount the camera at an angle that can see the baby. This robot is too short to do that it seems too. Same with feeding the fish if theyāre in a tank instead of a pond, unless it can scale up things. This is what I mean by I donāt know if the vision is fully flushed out yet.
4 Likes
My apologies for violating rules 
but I didnāt resist to write what I written
2 Likes
Regardless, I think the Purism technical model reduces the legal and ethical issues. As some technology moves in the direction of greater autonomous operation, it does raise those issues. Who is liable, who is responsible, when something goes wrong or is intentionally bad? The manufacturer? The vendor? The owner? Obviously this will vary from jurisdiction to jurisdiction, and I donāt think these issues have really been well-tested in court in most countries anyway.
With MiMi, you are the operator and hence you are most likely liable and responsible for what you direct MiMi to do.
1 Like
With the Librem 5, I saw the value in such a phone before I even discovered that a company named Purism was selling such a phone. So making a pre-order for such a phone and taking such a risk back then seemed reasonable as an early adopter.
Personally, I was unhappy with how Purism managed that project and my money. I eventually did get my Librem 5. The phone is good. But I was still disappointed enough to decide to never buy anything from Purism again until after development was complete and the product could ship to me right away. Once Purism has your money in a very loosely defined and non-specific contract, everything goes according to their terms, not necessarily what one may think is reasonable or fair to you. If youāre unhappy with how your money is managed after you send it to them, there is nothing you can do about it. If they want to wait twenty to thirty or more years to complete the project, you will wait that long. Any promise to you is very weak and undefined. Why not let them complete their robot product first, and then you can buy one if, at that point, you still want one. Life priorities and interests change for us over time. Many people changed their mind about the Librem 5, in just a few short years, and could not receive refunds, even though refunds were promised as a part of the original deal, simply by asking for the refund. As vague as that contract was, the refund promise was broken. Patterns are patterns. We know now that Purism will violate their agreements if they want to. At least now, they solved that issue of refunds by not offering them, up-front. Have you ever known anyone who had a stock investment that they couldnāt cash in, even at a loss, and that vested at an unknown time in the future? What if that same investment offered no financial return, only credit toward a product that you can buy later, at an unknown price? Just keep sending money and eventually, you will receive something (who knows what?) for it.
When Purism started the Librem 5 campaign, they had a history of shipping computer products. So it was reasonable to assume that Purism had the competencies needed to develop a phone. But Purism has not ever sold an AI or robot product. It could take decades to develop a first product. How long are you assuming that you will wait for a real product to arrive to you? Are you prepared to let them take 10x that long?
The only way I would take part in a Purism campaign now (as an informed Purism customer), is if a third party managed the money and if the delivery of a product was clearly defined with a hard delivery deadline. This would give Purism obligations to meet deadlines. Perhaps those deadlines should be generous in Purismās favor. But some hard delivery deadline should exist. If the deadlines are not met, then the third party that manages the money issues a refund check to you, regardless of how Purism is affected. A deal is a deal. To send money to Purism as some kind of pre-payment towards an undefined product at an unknown price, to be released at an unknown time in the future, is not a good business decision on the part of the buyer. If you have money to spare and want to gamble with it, thatās fine also.
3 Likes
Wouldnāt this essentially prevent any sort of crowd funding, because it would require Purism to put up the funds for the R&D given thereās always a chance that all crowdfunders might ask for a refund?
That said, MiMi brings to mind for me a question:
Does building the MiMi crowdfund project come at the expense of the much needed continuation of the Librem 5 project, or at the expense of Purism order times between when the order is sent and when they ship? My last two orders from Purism were 5 and 8 months ago, and it sounds like the order from 5 months ago is planned to start shipping next month. But the order from 8 months ago (sometime last year), I donāt really know, I imagine itās just stuck on parts maybe for a very long time.
Iām happy to be patient but those combined with continual claims that Crimson is not quite ready for L5 leave me with this feeling that there are just a lot of things to do. If I lost my job today, I could probably keep quite busy for a while trying to get my L5 to catch up with my ideals for what I wish it could do. I go birding with my parents and whenever they want to identify a bird song, an AI on their phone does it for them. When they want to track the birds they saw, their phone turns on a birding specific GPS history and records their sightings along with it.
And thatās just one of many situations where ā although Purism devices have made my life better by reducing corruption in the systems I interface with every day (i.e. less Google and Microsoft and others) ā these Purism devices constantly remind me of the need for me to go and either build or find online software solutions to my problems that in the mainstream surveillance capitalism society the people purport to have already solved. And often these arenāt even problems of hardware ā the Librem 5 could do all the stuff I want, if I had the right code ā but rather they are problems of software.
But between it all, the Purism stuff makes me think the future would be better if it had more evolution of improving the existing systems. And from a distance, before choosing to get invested or interested, the MiMi project looks like something tangential and not parallel with regards to improving these systems. I donāt mean to be a downer. But Iām not sure how else to say it. If AI robots are ubiquitous in 2040 and I am forced to have one, you can bet your patootie that I would rather have one from Purism than from a Facebook/Apple/Google/Microsoft/Amazon style corporation. But, in the short term, I want to make sure my existing Purism stuff can continue to evolve and doesnāt stagnate.
To be honest, I think I would sooner invest in a crowdfunding campaign for Super Duper Liberty Phone 8GB RAM with doubly fast processor and GLES3 that costs big overpriced $4000 price tag than in the extremely forward looking robot. I would like to find some libre AI system that can outsmart ChatGPT and run from the comfort of my own existing Purism hardware, but that doesnāt depend on someoneās silly nonfree AI model weights nor on Facebookās LLaMA or something. If I could prove to myself that these LLMs people keep touting are going to run handily on freedom respecting hardware, and not require some āsoftware as a serviceā bollocks, I could be more confident of my little AI robot 15 years from now in 2040.
And so in these combinations of my personal needs ā and some of them are my fault or just my own personal ignorance, but they nevertheless exist ā there are a lot of areas and directions that the experience of using my Purism hardware makes me wish we would go and evolve into. But almost nothing among them had made me picture a little MiMi Robot. Because until I get computing technology to where I can so easily wish it should be, where the ethics are really all there and not just a false sense of security propped up by a few people like me shoveling over our extra money or choosing to reject a social/technology landscape thatās obviously broken, itās hard for me to picture getting the full use out of a robot body.
Itās almost like, imagine one of the Agents from the Matrix movies standing there, and he looks down at MiMi and says, āBut Mr. Anderson, what good is a body if you areā¦ unableā¦ to think?ā
ā
And then it would be similar to the āunable to speakā scene in the original movies, but instead of the main characterās mouth melting away, in this case it would be MiMiās brain melting away into proprietary software running on proprietary nVidia cards that require nonfree drivers.
2 Likes
My point was simply not to give Purism a blank check when it comes to terms and deadlines. If a person wants to, they could expect an eventual refund after ten years if no product has shipped to them by a ten-year deadline (try selling that Purism). But expect some deadline. Donāt just leave everything open, expecting that Purism will meet your expectations, no matter how reasonable you guess they might be, when they never made that agreement with you. And when it comes to refunds, it should be only after Purism did not keep their agreements with you if those are the terms. How can you reasonably expect a refund when there was no agreements that can be broken anyway? But if there are delivery deadlines, there needs to be a promised refund if the deadline is not met, and you then need a third party to write the check and send that refund to you. We already know that Purism is unlikely to issue a refund, even if they agree in advance to give refunds (which they are not doing hereā¦ yet). Otherwise, donāt pretend that you are making a purchase when youāre really making a donation. This canāt even be called crowdsourcing, unless itās for pure charity. The early crowdsource participants for the Librem 5 were at least promised a phone. This robot fund literally promises you nothing what-so-ever except what Purism deems at some undefined time in the future, you should get. If youāve paid-in for decades and the robot costs $50K by then, and you have put in only $10K by then, then you would still need to pay another $40K to get anything (remember, no refunds). Maybe by then, youāre 80 years old (or dead by then) and Walmart will sell you a much better robot for only $200.00. Yeah, I get it. The Walmart one is not FOSS. Of course, I guess that changes everything in this scenario, right? (not so much).
2 Likes
That isnāt the point. You are not buying a robot. You are funding the development of a robot.
By definition therefore you are ceding much more control over how and where that development goes and how long it takes and, yes, there are risks (some of which are spelled out in the information provided by Purism).
The Ts and Cs are clear: There are no frigging refunds.
If you arenāt prepared to accept the risks and the Ts and Cs then noone is forcing you to fund this project.
Noting though that that would be your choice to do so. The Ts and Cs spell out that you can stop paying at any time. That is, you can cap your financial exposure.
Your minimum financial exposure is $5. You can think that itās a worthwhile idea, put in $5, walk away, leaving Purism to it, and come back in however many years time if and when a product is available for sale and decide whether you want to buy the robot, given the features and specification available and the asking price.
I donāt think that putting in $10K over ādecadesā would be very sensible unless you are quite wealthy - but if someone is quite wealthy then he/she/they is free to devote money in whatever way is deemed appropriate by the person in question.
Indeed. If you invite a device into your home, particularly if it has a mass of sensors and radios, then you are taking a big risk and, worse still, itās blackbox so you canāt even evaluate that risk before purchase.
2 Likes
Iāve noticed that the job postings donāt have compensation ranges posted, which are required by Washington state law (Assuming the SPC filing means Purism is operating from Washington; if itās California, they also have pay transparency laws (See āWhat is Californiaās salary range disclosure law?ā)). Please correct me if Iām wrong here, though I do think it would be a nice gesture even if itās not required by law.
2 Likes
The contract was indeed broken once, but this is definitely not a pattern. It may never happen again, especially since refunds arenāt promised anymore.
2 Likes
@JCS I should point out that the counter in regard to " Total and Goal (Updated Daily) " is not being updated daily. Purism ā MiMi a Crowdfunding Campaign to Make a Humanoid Robot
3 Likes