The ASR and NLU are running on a Raspberry Pi 3, and best of all it is free and we are open-sourcing it, starting with the NLU https://github.com/snipsco/snips-nlu
https://medium.com/snips-ai/snips-nlu-is-an-open-source-priv...
The whole platform runs on-device which makes it ideal for privacy, cost, and to allow it to run when there is no network
We are available in English, French, German, and soon Japanese and Korean, with more European languages coming!
We would love to see what you can build with it, and linking it to Ikea lights should be 10 lines of Python
Take a look at what some people have built with it: https://github.com/snipsco/awesome-snips
and a few tutorials to get you started: https://medium.com/snips-ai/building-a-voice-controlled-home...
3rd party services going away is a legitimate concern, but that's true of basically every piece of technology these days.
Someday someone might come along and build the thing you want. Perhaps they will build on this project.
Now, if there is a flaw in the criticism, address that, not the mere fact that it is criticism.
People should simply realize the product isn’t for them.
Often in long conversations, people chime in without reading the conversation and several people make the same “observation”.
In this conversation, we know there are downsides to cloud-based voice recognition. Telling someone who built it that it’s not for you, doesn’t add any value.
Realistically everything in this hypothetical conversation will be things that “everyone” knows. Unless you’re discussing confidential info about EVs, everything is by definition public knowledge and probably even common knowledge. Most conversations are about opinions rather than facts.
Is the guy who comments about EVs having all their torque available from a standstill also adding nothing to the conversation? What about the guy who mentions the environmental friendliness? Or the one who mentions the environmental costs of lithium mining? Who is really adding to the conversation here?
> In this conversation, we know there are downsides to cloud-based voice recognition. Telling someone who built it that it’s not for you, doesn’t add any value.
And yet it’s the top comment and the top reply provides info about a platform that doesn’t have the same drawback. I think this is pretty clearly value added.
So, my point about someone drowning out someone’s project is proven by the top-rated comment about how it doesn’t work for them.
Putting aside Snips for a moment, my point with my last comment was to note that you took the “upvotes aren’t a meaningful indicator” tangent and ignored the rest of the reply. You ignored the fact that the criticism led to a meaningful reply and did in fact contribute to the conversation. You also ignored my questions about what constitutes actual contributions to a conversation. Is it just positivity? Because most of the positive stuff people post is still not new and it seems unreasonable to me to claim that criticism is at heart not contributing to the conversation while fanboyism is.
Hopefully, both projects get a few more people involved.
I wasn't interested in a product that has to report to the NSA that I turned my lights on, and I didn't know about those...
So I'm happy that this comment thread is not the echo chamber you desire.
The primary benefit is that remote control can be effective from around the world and outer space.
This is countered by the reality that your command might get bounced off a satellite, even if you’re across the room from the device you’re controlling remotely.
The converse additional problem is that you could be standing next to a device on the north pole, and someone from antarctica could override you.
We haven’t become sophisticated enough about this level of remote control yet, to mitigate these undesirable side effects of extended range remoting via internet tunnels.
There's a difference between core and nice to have services (eg, ability to tell you the weather), but I consider light, door, and temperature control to be "core": having a dozen separate points of failure is just bad design. You could consider voice control nice-to-have, but if it works well, I have a feeling it would move to being core. If it doesn't work well enough to be used all the time (and thus irritatingly missed when it's broken), why bother?
I'm aware of Mycroft; anyone have experience with it or alternatives?