Can Josh Ai ask questions and respond?

[cybuch]

I am looking to replace my voice pods, but what is keeping me with VP, is it offline and it will ask me question and I can answer and then it will react.  Example, when I hit good morning button, if the temp is within a specific range, VP will announce the outside temp on my patio and ask if I want to open the back door, if I say "yes" it will open the back door.  Or if I hit watch movie in my living room, VP will ask if I want to close shading...etc.  

[l3anks]

Clarifying question : please define what is offline?

are you saying if connectivity is offline it asks you additional questions?

 

generally it won’t respond with questions to voice commands. The dealer will configure it so when you say “watch movie” it will close the shades, dim the lights, etc 

[RAV]

Josh is an input command system, but will perform different actions based on status of devices, time, weather.

Your Goodmorning, could open the doors if the temp is in range and sunny, but Josh won't ask for a confirmation.
Or Josh can start the movie, and if it's day, lower the shades, but if it's night leave them open, etc.

Also, you said button. Josh primary input is voice (or it's app), but you can send it commands from Control4 to match scenes in Josh that you've set up, and Control4 can receive commands from Josh through the Josh driver, 999 trigger inputs.

[cybuch]

23 hours ago, l3anks said:

Clarifying question : please define what is offline?

are you saying if connectivity is offline it asks you additional questions?

 

generally it won’t respond with questions to voice commands. The dealer will configure it so when you say “watch movie” it will close the shades, dim the lights, etc 

No, I mean it doesn't require an internet connection.  Thats not a deal breaker though.

 

9 hours ago, RAV said:

Josh is an input command system, but will perform different actions based on status of devices, time, weather.

Your Goodmorning, could open the doors if the temp is in range and sunny, but Josh won't ask for a confirmation.
Or Josh can start the movie, and if it's day, lower the shades, but if it's night leave them open, etc.

Also, you said button. Josh primary input is voice (or it's app), but you can send it commands from Control4 to match scenes in Josh that you've set up, and Control4 can receive commands from Josh through the Josh driver, 999 trigger inputs.

I mean, can I invoke it from a button press.  Clearly it is a one way system, you tell it what you want it to do.  I just wish someone would come up with a good voice pod replacement.

[RAV]

13 hours ago, cybuch said:

I mean, can I invoke it from a button press.  Clearly it is a one way system, you tell it what you want it to do.  I just wish someone would come up with a good voice pod replacement.

You can press a Control4 button, and have Control4 send a command to Josh AI to invoke a scene. (scenes in Josh are more like macros, its what they call the programming portion of the AI) That scene could be nothing more than a voice response, or invoke programming on the Josh side, although most typically that is a duplicate area of what can already be done in composer.

I expect the feature of Josh allowing secondary voice input after a trigger, will come shortly. That was the progression of Alexa as well. Josh AI is an evolving system.

[l3anks]

15 hours ago, cybuch said:

I mean, can I invoke it from a button press.  Clearly it is a one way system, you tell it what you want it to do.  I just wish someone would come up with a good voice pod replacement.

You can program a single, double, triple tap on the micros themselves to perform an action 

[nwanerka]

So I know this is a newbie question,  but what is the big difference between josh.ai and Alexa and Google?   So if I have a customer that says, why spend all the extra money on josh.ai, what can I tell them?

[lippavisual]

Josh doesn’t sell your personal info/ search history. 

[RAV]

It's a totally different experience.

Imagine you're a CEO. You want coffee. You ask your personal assistant for a coffee. They know size, where, add ins, etc. It's just done. Dave, please get me a coffee. Poof, magic.
Instead your assistant is out sick, and the temp Alexa is going to get your coffee. You need to provide every detail, exactly as the temp expects so they can get it right,if you make one mistake, leave out a detail, it's your fault, not the temps, Alexa makes you work for her. Even after she gets the coffee, she's doesn't know where to place it, you need to tell here that too.
No big deal, it's just one day. But your assistant is out for a week, and EVERY day, you have to detail your coffee order again and again, Alexa never learns. If your request doesn't match, no actions. Alexa is neat at first, but can wear off, or you need to spend more time making routines and learning the way she want's you to speak. It's work.

Josh is natural language, no meeting the exact syntax. So it's easier and more reliable. It knows alias to rooms: Gym, Exercise, Fitness etc. so everyone can use easily. It remembers usage, so things like 'turn them off' works. It's faster than Alexa. And it's focused on control, so it can be smarter in processing commands. AI isn't a cute tag on to their name, Josh is intelligent.

Josh does more music integration, especially with Sonos. It's also end user tweakable, they can easily make their own scenes. And it's faster than an app, or a touchscreen, how many button presses does it take you to play a song on demand? 7? 13? you'd be surprised how many if you count. And, it's also not another damn screen. All day long we're looking at screens. Work, phones, tablets, heck even the grocery store. I want a break, "Hey Josh, lets' unwind." (hopefully there's a bartender interface soon).

It's been said here many times, people are moving away from large control systems. Josh is an add on layer that makes it fun, and functions as whole home control bringing the pieces together. But their sandbox of products is small. You could build a very reliable, powerful system, with their sandbox of products: Lutron, Sonos, Ecobee, Samsung/Sony, Roku. But, with Control4, it brings in all the other pieces too: garage doors, locks, irrigation, IR and Serial control stuff, other shade vendors, and Control4 lighting integrates as well as Lutron. And voice isn't always the best interface, sometimes you want a touchscreen for parties, and viewing information.

And it's private, no one collecting all your data, and that's why you pay a license fee. $50 or $100 a month depending on size of project. What's your data worth, and a better experience too.

 

[malelan]

35 minutes ago, RAV said:

It's a totally different experience.

Imagine you're a CEO. You want coffee. You ask your personal assistant for a coffee. They know size, where, add ins, etc. It's just done. Dave, please get me a coffee. Poof, magic.
Instead your assistant is out sick, and the temp Alexa is going to get your coffee. You need to provide every detail, exactly as the temp expects so they can get it right,if you make one mistake, leave out a detail, it's your fault, not the temps, Alexa makes you work for her. Even after she gets the coffee, she's doesn't know where to place it, you need to tell here that too.
No big deal, it's just one day. But your assistant is out for a week, and EVERY day, you have to detail your coffee order again and again, Alexa never learns. If your request doesn't match, no actions. Alexa is neat at first, but can wear off, or you need to spend more time making routines and learning the way she want's you to speak. It's work.

Josh is natural language, no meeting the exact syntax. So it's easier and more reliable. It knows alias to rooms: Gym, Exercise, Fitness etc. so everyone can use easily. It remembers usage, so things like 'turn them off' works. It's faster than Alexa. And it's focused on control, so it can be smarter in processing commands. AI isn't a cute tag on to their name, Josh is intelligent.

Josh does more music integration, especially with Sonos. It's also end user tweakable, they can easily make their own scenes. And it's faster than an app, or a touchscreen, how many button presses does it take you to play a song on demand? 7? 13? you'd be surprised how many if you count. And, it's also not another damn screen. All day long we're looking at screens. Work, phones, tablets, heck even the grocery store. I want a break, "Hey Josh, lets' unwind." (hopefully there's a bartender interface soon).

It's been said here many times, people are moving away from large control systems. Josh is an add on layer that makes it fun, and functions as whole home control bringing the pieces together. But their sandbox of products is small. You could build a very reliable, powerful system, with their sandbox of products: Lutron, Sonos, Ecobee, Samsung/Sony, Roku. But, with Control4, it brings in all the other pieces too: garage doors, locks, irrigation, IR and Serial control stuff, other shade vendors, and Control4 lighting integrates as well as Lutron. And voice isn't always the best interface, sometimes you want a touchscreen for parties, and viewing information.

And it's private, no one collecting all your data, and that's why you pay a license fee. $50 or $100 a month depending on size of project. What's your data worth, and a better experience too.

 

this guy sells josh.ai..

[JSTRONG]

21 hours ago, malelan said:

this guy sells josh.ai..

He sure does. He understands what josh is. A control system unlike Alexa which is a personal assistant with some voice control abilities. And josh has some really good stuff in the works. 

[Control4Savant]

https://youtu.be/exU3worNLck 

Feel free to watch the keynote from the Micro and hardware release. 

[Neo1738]

Honestly all of this has me really tempted to make the leap....just have to convince the wife!

[cnicholson]

Back to the OP's question, If you wanted to simulate "Josh asking me a question" (at least simple Yes/No questions), I assume you could tell Josh to speak arbitrary text (to "ask" the "question").   The issue is whether you can hardcode the "context" and have Control4/Josh take the appropriate action.

Example:

1.  Your "Good Morning" event could do what it normally does and, as final step

2.  have Josh say the arbitrary text "Would you like to open the door?"  (Josh doesn't "know" it asked a question)

3.  You say to Josh "OK Josh, Yes," can you have the system

4.  know the state/context and respond accordingly.  (i.e., interpret the "OK, Josh yes" (in specific context) as meaning the same as "OK Josh, open the [appropriate] door."

This all seems plausible-- but I don't know how easy/hard it would be to implement reliably.  For a small handful of special cases, it seems doable.  But is it?

[malelan]

as far as i'm aware this is not possible today.

you can force josh to say things, but he isn't going to be aware of the differences with responses.