Funding Friday: An Open Source Voice Assistant

I backed this project last week. I think open source and transparency is badly needed in this market.

#crowdfunding#voice interfaces

Comments (Archived):

  1. Rob Underwood

    This is great! I was spending the morning preparing for our next Open Source Fin Tech Meetup coming up on the 12th; seeing some more cool open source tech is a boost. Totally agree that “open source and transparency is badly needed in this market”.

    1. DJL

      I’m actually interested in the underlying AI driving the software. We looked at some open source “virtual assistant” software to incorporate into our platform.

      1. P Donohue

        Everyone talks about open source, like it is good. I don’t know anymore. Tech has a massive security problem. I you have the funds hire an ethical hacker to go after everything you have. You may be shocked at what you (they) may discover.

        1. DJL

          Being in cyber security for a living – I tend to agree! Look at the holes in Mozilla versus Chrome. Nothing like a possible loss of money to get companies to fix stuff!

          1. P Donohue

            Regarding security, DOD+ looks at Kaspersky as a potential problem, thus no DOD+ machines have it nor run it.So, I have been wondering about PyCharm. Does it has a Kaspersky problem? Has anyone looked into that? Seems like a great platform to me.

          2. DJL

            I have never used that platform. But since it was developed overseas, you have to make the default assumption that it might be compromised. If I wanted to spy, I would create an innocuous plugin to an open source project and let er rip.

  2. Frank W. Miller

    I totally agree with you on supporting the open source here.I don’t know why people buy these things. I have too many people talking at me already. The last thing I want is my machines talking to me…

    1. jason wright

      should Mycroft bypass the hardware and team up with open source mobile initiatives (Copperhead, Lineage, UBPorts, Sailfish, Fairphone 2 OOS), or does it need the sales (revenues)?

    2. falicon

      These smart speakers are all the rage right now…it’s cute and fun to “Hey Google” right now in these early days…but just wait ten years or so, when these things are old enough to have teenage attitudes…suddenly it’s going to respond with stuff like “Hey yourself” or “why do you want to know that!?” or just a long sigh before ignoring me…….and you know it’s only a matter of time before the situation is flipped on us…and instead of us asking it to do things…it’s going to be demanding us to do things…”Hey Kevin, wash the dishes”, “Hey Kevin, change my batteries”, “Hey Kevin, lose some weight”… to which I say, “Hey smart speaker, go F yourself!”:-)p.s. comedy bit; clearly still a work in progress…

      1. Michael Elling

        Sprint’s superbowl ad was the clear winner imo.

    3. P Donohue

      I was one of the first people to sell cell phones in 2nd tier California market. For year 1,2,and 3 PacTel recognized us as having the most activations in their market.To a word, that was the first comment I encountered after my pitch. Today, if you tried to take the phones away from those same people, you would come away bloody. Go figure.

      1. Frank W. Miller

        Not me. Please take my phone, please! 😉

  3. kenberger

    I’m somehow reminded of DuckDuckGo and of Boxee.

  4. gorbachev

    The device needs an account at the service creator. Once they go belly up, you’ll have yet another useless IoT device filling your e-waste drawer.

    1. pointsnfigures

      I don’t know about that : )

  5. Kirsten Lambertsen

    Own a Mark I and ordered my Mark II after seeing Albert’s post. There’s something fun and Jetsons about Mark I’s form (while not as practical as II).They probably need to think about doing a bounty campaign to get more skills created. But everyone who believes in decentralization and open source should really be jumping in to drive this project forward in some small way.I also learned about Project Voice from the MyKroft folks:https://voice.mozilla.org/Everyone go contribute 5 minutes of your ears and voice 🙂 They need real-world recordings (you know, with the TV playing in the background and the dog barking). I’ve recorded 100 statements, so far. It’s fun.

    1. Vasudev Ram

      >There’s something fun and JetsonsHaha, as a kid, I once dreamed up an invention (partly triggered by living in a multi-storey apartment building at the time) that I called the Aircar. The idea was that it would be a small vehicle (I imagined it as something like a submarine / kayak sort of thing that you sit in, the vehicle’s walls come up to about your waist or a bit higher, and it is like a VTOL (Vertical Take Off and Landing) aircraft, so I could get into it (while it was parked on the floor of my apartment, in a room with a window, start the engine (jet propulsion, natch), make it rise up vertically, then fly out through the window, go around town, do my stuff like shopping, visiting friends, whatever, then fly back home, enter through the window, and land on the floor. Went around telling my friends about the idea.Later on I realized that it must have been inspired by (in fact, pretty much the same as) the vehicle in the Jetsons comics that I used to read, and even later, I read about the Moller M400 Skycar:https://en.wikipedia.org/wi

  6. TeddyBeingTeddy

    Is there an app that allows me to listen to text on my mobile screen? A “listen to text” button, reading “aloud” emails, article and books etc? I’m sick of reading and want to listen instead.

  7. Vendita Auto

    Mycroft. Sherlock Holmes brother good DNA

  8. William Mougayar

    So this is like Alexa or Echo, but where “openness” is driven by the market vs. by Amazon or Google?It reminded me of this other similar Kickstarter project from about 6 years ago. Same vision, but I was too early. https://www.kickstarter.com

    1. JamesHRH

      its driven by Fred not being down w Amzn & Goog keeping all that tasty data for themselves.

      1. sigmaalgebra

        It should be realized, of course, that nearly all the currently common cases (awful alliteration) of “that tasty data” is, for mining useful information, a mountain range of great variety of low grade ore and usually more frustrating than “tasty”.

  9. Vasudev Ram

    These two posts on my blog may be of interest to anyone who wants to play around with speech tech in Python – both ways – speech recognition and speech synthesis:Speech synthesis in Python with pyttsx:https://jugad2.blogspot.in/…Speech recognition with the Python “speech” module:https://jugad2.blogspot.in/…Beginner stuff, but fun.

  10. JamesHRH

    Duck Duck Hello.

  11. sigmaalgebra

    There’s a standard severe problem with technology and, really, with any new product:The solution is too often still looking for a suitable problem.This situation is inevitable:(1) Since people will buy even a pet rock or have even just trivial stuff go viral, even for something silly there may be a market for awhile.(2) There’s (A) a huge mountainous continent of results in math, physical, social, and medical science, engineering, and technology available for solving problems and (B) a lot of problems to be solved.So, maybe pick a pair, one from (A) and one from (B) and go for it!!!!!No, really don’t: Instead be careful!First criterion: Have some good reason to believe that the first good or a much better solution will be a must have for enough revenue to make a successful business.For just part of the challenge, in a little more detail, a friend once explained to me his view of the challenge:There are lots of solutions lots of people would like to have, but those solutions cost too much. There are lots of cheap solutions, but nearly no one wants them. For me, no way do I want a “voice assistant”.For the simple uses in the video clip, I already have good solutions.Then for the goal of using artificial intelligence to carry on a real conversation, no thanks:First, I see nothing in the world of AI that would make that possible.Second, even if such a thing in some simple cases seemed to be possible, the resulting conversation would be so empty it would be insulting. Too soon too many users would figure out that they had been talking only to a silly, dumb machine and wasting time, had been insulted, had gotten frustrated and angry, and had challenged the world record for distance of an object thrown by a right arm by a human.With this idea of a stand alone voice assistant, there is a lot missing here:E.g., (A) on my computer, I have a lot of information, and from some simple means it is well enough organized that I can find nearly all that information quickly. (B) As I go along, I collect more information and then for associated URLs, file tree names, etc. add to my collection of information.Right, the information is in a taxonomic hierarchy using the Windows NTFS file system! And I have some other DIY solutions I like!So, from 20,000 feet up, I am just one person, have just one collection of information, it’s on my computer, and I and use that information and add to it hour by hour daily. So, in particular, I don’t want separate, separated silos of information.But, if a voice assistant can catch on, then my project with some significant, unique utility for nearly everyone on the Internet looks still better!

  12. falicon

    Duplicating my comment from when Albert mentioned this project (because I am interested in what the community here thinks about it all too):Voice is a very exciting space right now!It really is a new OS war….so it will be interesting to watch this approach over the next year or two and see how it stacks up against Alexa, Google Home, and now the Homepod from Apple too. Will it be the Linux of this space or will it gather enough consumer support to become *the* standard?I’ve got apps built for the Alexa and the Google Home…will prob. get something onto the Homepod before too long…so I keep a fairly close eye on what the ecosystem and dev community around all of this is doing/into.So far, Alexa is running away with it all (even though the dev. community is *very* frustrated with the current monetization options and struggles mightily with discovery and user retention — but it’s where the users and most of the attention is).Google is starting to make a much stronger approach (and both Alexa and Google are starting to make some serious biz. dev deals/pushes under the radar right now — the ad dollars are really starting to flow in).Apple will get it’s usual chunk because of brand loyalty and how tightly integrated their device will be with the Apple ecosystem of devices (specifically the Apple watch and Apple tv to start).So a third party play has A LOT of challenges to overcome…but also has a lot of freedom to innovate on the small things that consumers and devs actually want (and care about — like the multiple names and faces, while a small technical thing, adds a lot of personality and fun to the overall device).Anyway – super fun and interesting space to be watching and or a part of right now.Disclosure: My startup, Veritonic, is the audio effectiveness platform helping brands and publishers measure and understand the value of Audio across all of their marketing. We are currently doing a lot of work with companies to help them figure out and understand what voices and sounds work best for their brand/needs across these devices. – you can learn more at http://veritonic.com

    1. Twain Twain

      Exciting but no cigar. Serious foundational problems with natural language frameworks.* https://www.inc.com/sonya-m…* https://www.wired.com/story…Sept 2017 research said only 6% of users remained active on Alexa app in the second week after downloading it.Amazon just invested another $3.5 million in their university challenge for teams to build better Alexa HCI.They also invested in Pulse Labs startup to do more quality testing of user voice interactions:* https://www.fastcodesign.co

  13. Michael Elling

    This is a good first step, but where is the exchange? Where are the settlements that provide incentives and disincentives? The shared benefit model only goes so far.