'Mycroft' Open-Source Voice Assistant Out of Funds, Can't Fulfill Remaining Kickstarter Rewards (kickstarter.com) 46
In 2019 Slashdot covered Mycroft, an open-source voice assistant for Linux-based devices (including Raspberry Pi boards). But this week the company's CEO posted on Kickstarter that "without immediate new investment, we will have to cease development by the end of the month....
"We will still be shipping all orders that are made through the Mycroft website, because these sales directly cover the costs of producing and shipping the products. However we do not have the funds to continue fulfilling rewards from this crowdfunding campaign, or to even continue meaningful operations."
The announcement details Mycroft's long, strange trip, from a hardware-focused partner that couldn't provide stable hardware to their switch to using off-the-shelf parts — followed by supply chain disruptions (with hefty import and manufacturing fees): The best plan we could devise to fulfill the remaining campaign rewards was to use the slim margins we have on new sales to cover the increased costs of hardware production. With that plan in mind, we pushed forward and started production. We got plastic injection molds cast. We started printing custom PCBs. We engaged audio engineers to optimize the quality and volume of the sound output. We got the device FCC and CE approved. Many of these steps took multiple iterations to get right, and there are many more things that I'm glossing over. All up this costs — a lot of money. Far more than the total contributions from the campaign, which is why I personally committed so much additional funding. I could see a clear way forward that strengthened Mycroft as a project, as a business, and as a community.
So what went wrong? The single most expensive item that I could not predict was our ongoing litigation against the non-practicing patent entity that has never stopped trying to destroy us. If we had that million dollars we would be in a very different state right now.
With so much of our focus on hardware, and less funding to devote to improving our software — the quality and features available on the Mark II at launch were clearly underwhelming. It is more robust and stable than it has ever been, but this came at the cost of fewer new features. That in turn I believe has resulted in less than flattering reviews, and little mainstream coverage. The hardware itself has proven itself to be a solid base to work from, but without good reviews you get less sales, and without strong sales, the plan doesn't work.
Thanks to stx23 (Slashdot reader #14,942) for sharing the news.
"We will still be shipping all orders that are made through the Mycroft website, because these sales directly cover the costs of producing and shipping the products. However we do not have the funds to continue fulfilling rewards from this crowdfunding campaign, or to even continue meaningful operations."
The announcement details Mycroft's long, strange trip, from a hardware-focused partner that couldn't provide stable hardware to their switch to using off-the-shelf parts — followed by supply chain disruptions (with hefty import and manufacturing fees): The best plan we could devise to fulfill the remaining campaign rewards was to use the slim margins we have on new sales to cover the increased costs of hardware production. With that plan in mind, we pushed forward and started production. We got plastic injection molds cast. We started printing custom PCBs. We engaged audio engineers to optimize the quality and volume of the sound output. We got the device FCC and CE approved. Many of these steps took multiple iterations to get right, and there are many more things that I'm glossing over. All up this costs — a lot of money. Far more than the total contributions from the campaign, which is why I personally committed so much additional funding. I could see a clear way forward that strengthened Mycroft as a project, as a business, and as a community.
So what went wrong? The single most expensive item that I could not predict was our ongoing litigation against the non-practicing patent entity that has never stopped trying to destroy us. If we had that million dollars we would be in a very different state right now.
With so much of our focus on hardware, and less funding to devote to improving our software — the quality and features available on the Mark II at launch were clearly underwhelming. It is more robust and stable than it has ever been, but this came at the cost of fewer new features. That in turn I believe has resulted in less than flattering reviews, and little mainstream coverage. The hardware itself has proven itself to be a solid base to work from, but without good reviews you get less sales, and without strong sales, the plan doesn't work.
Thanks to stx23 (Slashdot reader #14,942) for sharing the news.
I want Open Source to beat spying Google/MS/etc. (Score:2)
But for some reason the spying bastards can make stuff that sells and makes profits, and the Open Source people
Re: (Score:3)
Re: (Score:3)
You missed mentioning George Soros. Gotta layer that on if you want to be acceptable to your target demographic.
Re: (Score:2)
I want a good, open, non-spying Phone, Tablet, Home Controller, Voice Assistant, etc. But for some reason the spying bastards can make stuff that sells and makes profits, and the Open Source people ... not so much.
The "spying bastards" have deep pockets to thwart patent trolls. Open Source can't afford protection from those scum-sucking bottom-feeding parasites. The solution is patent reform, and more broadly, IP legislation reform. But that won't happen, because of bri... er, lobbying.
Shakespeare wrote: The first thing we do... (Score:5, Informative)
Re: Shakespeare wrote: The first thing we do... (Score:2)
Re: Shakespeare wrote: The first thing we do... (Score:2, Interesting)
Re: (Score:2)
Well the other being it takes a lot of hardware to make these things work properly. Anti-patent trolling is just a side-effect.
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Can't they at least name that NPE? So we can check the details of the case and if at least their parents were invalidated.
I'm guessing you're not quite as motivated as your question suggests. I'd never heard of Mycroft, and yet I found the answer in seconds.
OP was asking for the NPE, not for Mycroft.
Disappointed to see it end like this (Score:5, Interesting)
I've been wanting a unified piece of hardware like this for a long time to muck around with the software side of a voice assistant. I haven't had time to futz with hardware so it would help to get solid pieces assembled into a good speaker.
Admittedly I'm part of the problem - even at the $350 price point before the price was raised further, I just felt too hesitant about long term support to bite. It's absolutely true that the downward spiral in the features kept me from investing it it.
Apparently these patent troubles were bad enough they wrote a children's book about it?! https://mycroft.ai/product/myc... [mycroft.ai]
Re:Another "failed" project (Score:4, Informative)
For once its not the fault of the company doing the Kickstarter project. They had to blow a lot of their limited funds fighting a scumbag patent troll. Its past time to reform the patent system so as to end patent trolls (won't happen though, not as long as the lobby group for patent attorneys continues to make "campaign contributions" to the politicians in Washington to keep the gravy train going)
Just $1 million? (Score:5, Insightful)
A million dollars seems like a lot to most of us. But in terms of building a voice assistant, it's next to nothing.
Making a computer voice speak intelligibly, and understand your words, is by far the easiest part. It's certainly not easy, especially when you start to try to support heavy accents or different languages. But still, the easiest part.
The hard part is understanding context and intent. When I ask Google Assistant, "What is the temperature?" it has to understand that I'm talking about the temperature outside, in my location, and it has to actually know my location. This is a trivial example. "How long will it take me to get to work?" similarly requires a lot of context. Every question or command requires lots of unspoken, but important, additional data, in order to provide a useful response.
Understanding this context requires a LOT of research and development time and testing, far, far more than a mere million dollars.
Re:Just $1 million? (Score:5, Interesting)
Making a computer voice speak intelligibly, and understand your words, is by far the easiest part. It's certainly not easy, especially when you start to try to support heavy accents or different languages. But still, the easiest part.
The hard part is understanding context and intent. When I ask Google Assistant, "What is the temperature?" it has to understand that I'm talking about the temperature outside, in my location, and it has to actually know my location. This is a trivial example. "How long will it take me to get to work?" similarly requires a lot of context. Every question or command requires lots of unspoken, but important, additional data, in order to provide a useful response.
Understanding this context requires a LOT of research and development time and testing, far, far more than a mere million dollars.
A generation ago I could talk to my mobile phone and ask it a few questions. It would do a few tasks for me like respond to simple questions, run programs and place voice calls. The hardware was orders of magnitude less capable than even the lowest end smartphones available today yet it did everything I wanted it to do mostly over BT when I didn't have access to a screen or normal input.
Today it is impossible to have even that same basic functionality without it being routed thru some creeptastic Google/Apple/Amazon/ad nausea spyware. Do I really care some fancy new assistant driven by a state of the art NLP model is able to alert me when the next planetary alignment of Venus, Mars and Jupiter is?
One does not need to spend billions in R&D to get the weather, make calls, send texts, set timers and alarms, change channels, search media libraries...etc..etc..etc.
Everything having to do with voice recognition over the 20 years has been subject of predatory legal shenanigans that has completely destroyed the industry. Mycroft is what happens to everyone who tries that does not have billions of dollars in the bank.
Re: (Score:2)
Your desires for what the voice assistant should or can do, doesn't reflect what most customers want. What most people want, is a magician that can do anything.
Re: (Score:2)
Your desires for what the voice assistant should or can do, doesn't reflect what most customers want. What most people want, is a magician that can do anything.
How do you know what "most customers" want?
Re: (Score:2)
How do you know what "most customers" want?
Two ways. I observe people, and I observe what businesses do.
Google, Microsoft, Amazon, and others, spend a ton of money--billions--developing their voice assistants. https://medium.datadriveninves... [datadriveninvestor.com] They wouldn't spend all this money on them, only to make them do things that people don't care about. Sure, they miss the mark sometimes, like Amazon making Alexa too much of a salesbot. But they do their homework.
Re: Just $1 million? (Score:2)
Re: (Score:2)
"a generation ago" I remeber those 2021 smartphones. Those were the days
Maybe not a full generation, but certainly half of one. The smartphone I had 10 years ago had tolerable voice recognition built in. And if GP was talking about hardware generations, it's been several.
Re: (Score:2)
Re: (Score:2)
The opposite of what you claim is true.
Initially voice assistants sent everything up to the cloud for processing. The raw audio was sent, they didn't even try to filter it on the device itself.
Nowadays Google does all the voice recognition on the device itself, at least for its own hardware. Even in airplane mode, speech recognition works and the phone can perform any task it can do locally, like setting an alarm or changing a setting. They did it partly because it makes the assistant more responsive, and p
Feature Not Bug (Score:2)
The Patent System crushed this startup, prevention meaningful competition with the wealthy and powerful entrenched interests.
Marking this bug report 'Works as Expected'.
Only Humanity could disagree.
How about starting simpler? (Score:2)
How about starting simpler? I just want a device that has a good mic, and good speech to text.. It can respond w/ different bongs and lights, no need for text to speech initially. They shot for the moon w/ displays, TTS, etc. Start simpler, w/ a good reliable device, earn trust that you can build things.
Website orders are still fulfilled (Score:2)
Or so they say [mycroft.ai]. Yes, it still sounds like they're nearing the end, but for now only the Kickstarter backers are out of luck, not so much those who ordered (or order) on the website. "All components have been purchased, and all Mark II orders outstanding and those placed in the future will be delivered."
What went wrong? (Score:2)
Nothing, really. The problem is that electronic personal assistants are largely useless: they are able to do very well what you can do just as well and without much effort, and incapable to do what one would really expect from a personal assistant to be able to do: you are not likely to ask your personal assistant to tell you what the tallest mountain on the planet is very often, but you might ask them to give you an intelligent summary of the news of the day, or deal with the nitty-gritty details of your v
Re: (Score:2)
Electronic assistants make technology more widely available by making it easier to use.
Re: (Score:2)
ChatGPT with a voice interface wrapped in a little device would probably sell well. Moreso if it could do things beyond answer.
KC company incorporated in Delaware.. (Score:1)
Re: (Score:2)
Re: (Score:1)
Re: (Score:1)
This is probably untrue, investors like Delaware corporations because they know the rules and how they will be enforced. Plus, states are free to charge a tax on out-of-state corporation doing business in that state (IIRC Mycroft HQ is in Hawaii).
It’s called the Delaware Loophole for a reason.