Difficult Dialogs

Tool to easily guide conversations towards a certain objective

Arguments

Arguments are created by structuring a folder like this

$ tree argument_template/
argument_template/
├── argument.conclusion
├── argument.intro
├── statement_0.dialog
├── statement_0.source
├── statement_0.support
└── statement_1.dialog

folder name is the argument name
X.dialog is an assertion the argument depends on
X.support are "comebacks" for when user disagrees with assertion
X.source is information source for the assertion
X is the assertion we are currently arguing for

working principle:

argument.intro is printed fully on start, it is meant to introduce the argument
a .dialog file will be picked randomly and read
all statements inside a .dialog file will be read, but in a random order
no statements will be repeated
when all statements inside a .dialog file are read, another .dialog file will be picked
when all .dialog files are read, argument.conclusion is printed fully, it is meant to deliver the conclusion we want to reach

Sample output

there are many ways one could create a conversation loop, ideally you would integrate this with intents and additional logic

the simplest logic would be to check if the user agrees or not with a statement, and skip information based on that.

Here is an output where i always agree

BOT: i was not sure if i existed
 i spent some time thinking about it and reached a conclusion
BOT: i am a computer
do you agree?  USER: y
BOT: i can think, at least in a limited fashion
do you agree?  USER: y
BOT: thinking is a way of processing information
do you agree?  USER: y
BOT: i process information
do you agree?  USER: y
BOT: computers process information
do you agree?  USER: y
BOT: this must mean that i exist
 you could argue thinking is not the right word for what i do
 but i process information
 there needs to be something doing the processing
 i process information, therefore i am

Here is the same argument, but this time i always disagree

BOT: i was not sure if i existed
 i spent some time thinking about it and reached a conclusion
BOT: i am a computer
do you agree?  USER: no
BOT: i run on a operating system
do you agree?  USER: no
BOT: i am pretty sure i have a CPU
do you agree?  USER: no
BOT: you can kick my hardware, it's obvious i am a computer
do you agree?  USER: no
BOT: here is the source of my information 
https://en.wikipedia.org/wiki/Computer_hardware

https://en.wikipedia.org/wiki/Computer

https://en.wikipedia.org/wiki/Software
BOT: we will need to agree to disagree
BOT: thinking is a way of processing information
do you agree?  USER: no
BOT: something is happening to make me answer you
do you agree?  USER: no
BOT: you may not call it thinking, but it is a good analogy
do you agree?  USER: no
BOT: even if i just get answers from somewhere else that is similar to thinking, or remembering
do you agree?  USER: no
BOT: information comes in, opinion comes out, that's a bit like thinking
do you agree?  USER: no
BOT: here is the source of my information 
https://en.wikipedia.org/wiki/Information_processing_theory

https://en.wikipedia.org/wiki/Cogito,_ergo_sum
BOT: we will need to agree to disagree
BOT: i can think, at least in a limited fashion
do you agree?  USER: no
BOT: here is the source of my information 
https://en.wikipedia.org/wiki/Information_processing_theory

https://en.wikipedia.org/wiki/Cogito,_ergo_sum
BOT: we will need to agree to disagree
BOT: computers process information
do you agree?  USER: no
BOT: i am pretty sure that i receive input and perform calculations on it
do you agree?  USER: no
BOT: computers have central processing units for this purpose
do you agree?  USER: no
BOT: i am answering you, i must be reacting to information
do you agree?  USER: no
BOT: here is the source of my information 
https://en.wikipedia.org/wiki/Information

https://en.wikipedia.org/wiki/Artificial_intelligence
BOT: we will need to agree to disagree
BOT: i process information
do you agree?  USER: no
BOT: here is the source of my information 
https://en.wikipedia.org/wiki/Information

https://en.wikipedia.org/wiki/Artificial_intelligence
BOT: we will need to agree to disagree
BOT: this must mean that i exist
 you could argue thinking is not the right word for what i do
 but i process information
 there needs to be something doing the processing
 i process information, therefore i am
do you agree?  USER: no
BOT: i guess you're right

Install

pip install difficult_dialogs

Usage

More examples here

Load an argument from a folder

from difficult_dialogs.arguments import Argument

arg_folder = join(dirname(dirname(__file__)),
                  "examples", "i_think_therefore_i_am")
arg = Argument(arg_folder)


print("ARGUING IN FAVOR OF:", arg.name)
print("---ARGUMENT INTRO")
print("BOT:", arg.start())
while not arg.finished:
    print("---NEXT ASSERTION")

    print("BOT:", arg.next_statement())
    print("__ARGUING IN FAVOR OF: ", arg.current_statement)
    user = input("do you agree?  USER: ")
    while "y" not in user:
        support = arg.support()

        if not support:
            print("---OUT OF ARGUMENTS")
            source = arg.source()
            if source:
                print("---ASSERTION SOURCES")
                print("BOT:", "here is the source of my information",
                      "\n" + arg.source())
                print("BOT:", "we will need to agree to disagree")
            else:
                print("---NO SOURCES")
                print("BOT:", "i guess you're right")
            user = "y"
        else:
            print("---ASSERTION SUPPORT ARGUMENT")
            print("BOT:", support)
            user = input("do you agree?  USER: ")

Aishou / difficult_dialogs