Chat bots lack sensitivity ?
The week has been stressful in a way, since there were submissions even from the other studio to do along with coding. My groupmate, Calvin and I, did a decent job at coding TALKO according to us. We had a good start, since we had already coded a bit of the Equipment section by ourselves. But then it was very robotic and not very conversational, so we decided to change it so the user feels more like its interacting with something which understands what he or she is saying and does not just give monotonous replies.
Although to think of it, these chat bots are coded to work the way they are. So technically, you’re just giving the user an illusion that they’re conversing to something that understands what they’re saying, while the bot is only actually catching onto words and replying. So isn’t that deceiving the user in a way? Then how does it make for good design? Unless we explain a human talking to another as picking onto words and responding accordingly too. As much as we say humans listen to another talk just so they can reply, humans have an additional factor of emotions playing in. Thus the reply is also based off that along with the words spoken. Chat bots lack that kind of sensitivity. Yes, AI could possibly be made to understand emotions but AI, according to me, shouldn’t be given the same amount of power that humans have. Just like Neil DeGrasse Tyson says, we don’t need to worry about these bots as long as there’s no emotions involved into them.
Coding for TALKO was good but had its ups and downs. Only when we started coding, did we realize the limitations of our concept. Concepts can be good solutions, but how realistic are they? Some things, can clearly be learnt only via making. I think the problems we encountered, were more often than not, minor errors within the code. Issues like indentation errors were getting annoying after a point, because we were coding all of this, and then it wouldn’t run because somewhere the spacing wasn’t right. And as much as this is not too big an error, the time constraints, the amount of work that needed to be covered up within that time frame, etc. was getting to us. Moreover, here and there we’d forget to put a closing syntax, and thus there’d be an issue with running the program.
There are still some here and there things we encounter, with respect to coding, that we’re unclear about. But we usually manage to make our way out. Moreover, I think I have learnt at this point that there’s more than one way to get the same output. So even if there’s one I’m not sure of, there’s always another way I can try doing the same thing. Similarly, there might be one way I know of getting an output, but there might be an easier one too.
Guess at the end, I can say I’m getting a grasp on it, and learning alternate ways to do things through coding itself. Also, looking forward to polish it further for A3, and work on a better conversation using TALKO.
