Bottie go get your gun

When thinking of chat bots, many people think of Siri, Alexa, or the little chat box down in the bottom of a website that most people avoid. When interacting with Siri, or Alexa, it is usually asking them to perform a task — play music, or search the web. I feel comfortable saying that most people do not go to these bots when they just need to chat with someone. I think this is because they sometimes get words wrong, and it is hard to talk to “someone” who does not understand you completely. Maybe that is because they do not understand the emotional baggage some of our words carry. The difference between those types of bots and chatbots on web services, or over the phone with choices already laid out for you is that there is not always the choice you would prefer. Sometimes it is quicker to just talk to a human about a technical issue you are having or reporting a claim, rather than going through all of the choice options until you find the one that somewhat relates to your issue.

For this project we were asked to construct a chatbot that chats about a controversial topic. I thought of doing gun control. I thought it would be a great topic because it is a very polarized topic, and I think that it is impossible to know all the facts. Also, most everyone has a side that they take. We really wanted to design our bot for that average user. An adult that can read, and maybe even reads the news occasionally. Our main goal was to just inform our user about the other side of the topic. This is such a polarized topic, that there is no way that we were going to change someone’s mind by just spitting facts at them. We hope to keep people educated about their decisions.

We understood that a chatbot is not the whole story. There was no way to create a chatbot in a week that can encompass everyone’s opinion and give them a very tailored response. We decided that we were just going to take the stance of pro- and anti-gun laws. Our team did individual research to make sure that we tried to get as many facts as we could. We looked at reasons people are pro- and anti-gun, and what factors are the things that cause people to pick a side. Using this information we created a basic chatbot that argued for less gun laws and tested it on users.

Classmates Testing our Prototype

The feedback was pretty average, not many people know how to criticize people without hurting their feelings, so a lot of people say our project is “good”. The best feedback was that the user was not sure how much his input mattered. I really respected this opinion, because he was right — his opinion really did not matter. Chatbots are hard to take into consideration all stances on the topic. He had to categorize his opinion in order to use our bot.

Taking the feedback, we then went back and added some features to our bot that we thought would help people feel like they had more control. I mostly worked on when people were interested in having less gun-restriction laws, and countering them. My goal was not not make them change their mind, but to educate.

I then had my roommate use the bot we created. She knew that it was preprogrammed, because the response time was so quickly, but she felt like her opinion was always heard (she leans towards having less gun restriction laws). I asked her if she felt like she had learned something, and she had.

Overall, I feel like our chatbot was persuasive, because it was not trying to change anyone’s mind. I think in order to actually change someone’s mind, you have to share stories with them, and not argue with them. You have to make them feel, and it is difficult to make a chatbot do that. I think that there are many ways that our bot could have been improved. It is hard to work on a chatbot with a team, because people have very different ideas on how to construct different aspects. It is very obvious in our bot that two people worked on pro-gun laws and the other two worked on anti-gun laws. If you were to have the same conversation with a person who was going to argue the opposing side, they would have a similar way of going about arguing their side.

Link to our video: