Code as a Language

During the course of my learning of Python, I realised that “ hackers and painters are among the most alike”- Paul Graham. Artists and Coders are both makers indeed. While painting, an artist fills his painting with all the thoughts and imaginations about what his painting should look like and so does a coder. Coder also fills his code with a deep thought about what his product should do and what his product should make the user feel when he runs the code!

Code is a way to make life systematic and automatic. I remember Harry Potter movies and how everything was magic, I feel today we are towards a similar world. The only difference between that world and this is, there was magic behind objects which made them automated, here code makes them systematic and automatic. In that case we are also living in a world which is magical. It is magical to see how things like NLP work on lines of understanding human language and how digital space is intervening in our lives. Such interventions are using “humane language” to intervene, which makes humans more responsive to digital products.

This is what we tried to do in our project on Conversational Agent. Our main task was to make it more humane and automated. Hence we started by mapping conversations of humans and recording their reaction on a particular sentence and their possible multiple replies. This led to building of structured conversation later. And this is how Interaction Design works- there needs to be a human touch to product to make people feel like interacting with them.

While working on project, I realised what does a prototype mean but more of- how much does it mean. A prototype is defined by its functionality. In the prototype user should be able to experience a mini version of the product and should display all the working features of actual program. Even in prototype, user experience matters which makes it necessary for prototype to be error free. While building a product it is very important to make prototypes and test it with user group. Sometimes, it may happen that feedback from user group or their response in interaction with product may lead you to redesign the meaning of your product. Hence user testing also becomes very important. After writing the code, I tried testing the code with people by giving minimum brief so that I can understand possibilities of human behaviours, which led to a better defined product in the end!

From the first stage of prototyping to last stage, we redefined it every time. We had to go back to the starting of conversation and think critically about each line our conversational agent has to say. We had to come up with a construction plan in which user is automated to reply accordingly. Basing our code on inputs, we could make it more user optimised.

Related posts:

When it comes to backend development in Python, Django is one of the best frameworks you can find. It enables you to do fast development because of the number of features it has and its built-in ORM…

It was a great summer at WCMA Digital. Summer means focused time to work on systems and infrastructure (yay!), without the delightful distraction of classes and students and faculty. It also means…

Alan Turing is responsible for engineering the Enigma machine, which decoded secret messages from Germany. In later life though he was a convicted criminal.