kriiilllee

Honorable
Nov 20, 2017
15
0
10,510
Hello peeps! This is probably not at all in the correct thread and im sorry in advance!
Im curious about this AI some people are creating, Its very similar to JARVIS in iron man, Ofc not as advanced as in the movie but still.
I was wondering if anyone here has done that and how its made?
The first video is very advanced and the 2nd video is more in the right lvl and that is what i wanna know how its made.
View: https://www.youtube.com/watch?v=vvimBPJ3XGQ

View: https://www.youtube.com/watch?v=oteZqZqvVQs
 

USAFRet

Titan
Moderator
"How its made?"
Some sensors (camera and mic), and some code to recognize preregistered commands. Or the ability to parse speech into text, and run a google search on that text.
An Alexa or Echo will do the same.

True "AI" is being able to ask "Why is the grass green?", and it can figure it out.
 

kriiilllee

Honorable
Nov 20, 2017
15
0
10,510
Okey cuz personally i would like to have something like the guy in the 2nd video but i dont know whats needed or how to create it to tinnish "product" more than the code but code should be able to be found on the internet right?
 

USAFRet

Titan
Moderator
I understand its custom but somewhere on the www there should be some code that u can use as ground and modifie along the way
"Speech recognition"
Then, parsing the resulting 'text', and deriving some keywords.

'tomorrow' + 'weather' + our preset location = query google (or whoever) for weather data pertaining to tomorrow.

'shopping' = Amazon
'Not Amazon' = Newegg

keywords, and the database to hold what should be done with a particular text or combination of words.

There is no generic thing you can just plug in and modify.

What is your coding background?
 
  • Like
Reactions: bit_user

bit_user

Titan
Ambassador
I understand its custom but somewhere on the www there should be some code that u can use as ground and modifie along the way
Here's a place to start:

https://fosspost.org/lists/open-source-speech-recognition-speech-to-text

I recommend seeing if you can get any speech recognition engine to work. Then, play with it to see what it does. After that, perhaps you can integrate it into a program that understands and performs a few predefined commands.

If you're asking how to build Alexa, Siri, Google Assistant, etc. you should understand that these all had probably hundreds of highly-educated professionals working on them for years. They're not the sort of thing you can build yourself, without at least a good understanding of computer science and AI.