Experimenting chatbots

Chatbots and AI have been around for a while but the fact that Facebook chose to launch this feature could mean that this technology is finally mature enough for mainstream adoption. Scott Milburn, a UX Designer alongwith his team in a recent hackathon, tried to build a Don’t-Worry-Mum bot and he gave his first hand experience in the article.

Don’t-Worry-Mum bot

The team decided to create a bot that would take care of mum’s nagging messages and keep her at bay  worry-free. Two developers built the framework for the bot. Scott Milburn and an other Designers determined how it interacts with the world. Facebook’s wit.ai was used to design and train the chatbot.

Designing the BOT

12 major scenarios ware taken as stories. The next step was to generate questions and answers for each of the 12 Stories. Questions were marked with an Intent and paired with a response from the bot.

Training the BOT

The bot was tested with more than 30 subjects. The results were logged in wit.ai, and reviewed. Whenever the bot was unsure of what to say, it provided the default response and logged the uncomprehended message in the Inbox.Training the bot was simply a matter of assigning Intents to the uncomprehended messages, a process wit.ai calls validation.

Bots make mistakes too

In the learning stage the bots can make mistakes. Next time when the bots encounter a similar message, it will give the correct response.

Chatbots are challenging to get right

Overall, chatbots are fun to build but difficult to get right. The conversational interface inherently lacks constraints, making it difficult to predict user behavior.

Conclusion

Chatbots have just raced off the starting blocks, and it will be a few more years before we know if they’re here to stay. As we leapfrog into the era of AI, we’ll probably see an increase in human-assisted chatbots with a gradual shift towards full automation.