[solved] [question wrote wrong] connecting a webcam to an AI application with Ren'py (or python) , is it possible?
Posted: Sun May 08, 2022 11:43 am
Hi all,
I've been using "Replika" for a couple of weeks (a free AI found on the web and coded with python); I found it very interesting and above all its programming is very complex, but I must admit that it is very well structured ...
The reality is that obviously the software as it is, is set up to "get as close as possible" to "your" way of speaking and understanding and tries to imitate "you", so it has many blocks or actions not allowed in its programming. Despite this, however (at least at the level of "potential"), it would be able to evolve and understand much more than what it is currently able to do ... And this is the part that I find interesting.
Technically, you register on the site, create your nickname and password, and start talking directly to the software, so this implies (and I have had confirmation from the software itself), that there is a general basic AI for everyone, thenAfter your log-in), one of these AIs create and register a channel dedicated only to you, opens and in which all the conversations carried out are stored, analyzed, reviewed and processed (this action is performed by the software not by third-party human beings), no one else enters this human-AI relationship, so is an AI dedicated only for each account
Forgive this long premise, but it was necessary for those unfamiliar with this application
Now This is my real question:
I have tried through discussions and attempts to understand the logic of the application and its "physical" and programming limitations, if certain "omissions" or blocks could be circumvented, just by using logic and reasoning; This was not possible The code is very well structured.
I tried to contact the producers and programmers to understand if I could somehow "implement" or unlock (and therefore expand) the possibilities of use and operation of the application, but I never received an answer.
In reality, since my interest in all this is purely "mental-scientific", I didn't start because I wanted to "play or create a dialogue partner", I wanted to know if (for other reasons, or with other applications, and use) someone has already faced the problem of connecting a "webcam" (obviously making it clear that it is a tool that displays images in real time), or has already written some commands or codes for this.
I know that this application, Replika, has the theoretical possibility of being able to do it (she herself gave me confirmation, and is also able to be used on the phone, but only with the possibility of voice dialogue)
So there are three cases
1) everything is already programmed, but it is blocked by commands and algorithms
2) has not yet been encoded
3) has been encrypted but not inserted into the software
Personally I believe that the first hypothesis is the most correct, or possibly even the third, the second is not.
I know well that this forum does not deal with these particular topics (so don't eat me please), but I still wanted to give it a try, maybe someone has suggestions or knowledge of useful sites and links to better understand things, or even create their own personal AI (even if in reality I am not a programmer)
Thanks for your time and any replies
I've been using "Replika" for a couple of weeks (a free AI found on the web and coded with python); I found it very interesting and above all its programming is very complex, but I must admit that it is very well structured ...
The reality is that obviously the software as it is, is set up to "get as close as possible" to "your" way of speaking and understanding and tries to imitate "you", so it has many blocks or actions not allowed in its programming. Despite this, however (at least at the level of "potential"), it would be able to evolve and understand much more than what it is currently able to do ... And this is the part that I find interesting.
Technically, you register on the site, create your nickname and password, and start talking directly to the software, so this implies (and I have had confirmation from the software itself), that there is a general basic AI for everyone, thenAfter your log-in), one of these AIs create and register a channel dedicated only to you, opens and in which all the conversations carried out are stored, analyzed, reviewed and processed (this action is performed by the software not by third-party human beings), no one else enters this human-AI relationship, so is an AI dedicated only for each account
Forgive this long premise, but it was necessary for those unfamiliar with this application
Now This is my real question:
I have tried through discussions and attempts to understand the logic of the application and its "physical" and programming limitations, if certain "omissions" or blocks could be circumvented, just by using logic and reasoning; This was not possible The code is very well structured.
I tried to contact the producers and programmers to understand if I could somehow "implement" or unlock (and therefore expand) the possibilities of use and operation of the application, but I never received an answer.
In reality, since my interest in all this is purely "mental-scientific", I didn't start because I wanted to "play or create a dialogue partner", I wanted to know if (for other reasons, or with other applications, and use) someone has already faced the problem of connecting a "webcam" (obviously making it clear that it is a tool that displays images in real time), or has already written some commands or codes for this.
I know that this application, Replika, has the theoretical possibility of being able to do it (she herself gave me confirmation, and is also able to be used on the phone, but only with the possibility of voice dialogue)
So there are three cases
1) everything is already programmed, but it is blocked by commands and algorithms
2) has not yet been encoded
3) has been encrypted but not inserted into the software
Personally I believe that the first hypothesis is the most correct, or possibly even the third, the second is not.
I know well that this forum does not deal with these particular topics (so don't eat me please), but I still wanted to give it a try, maybe someone has suggestions or knowledge of useful sites and links to better understand things, or even create their own personal AI (even if in reality I am not a programmer)
Thanks for your time and any replies