How does Artificial Intelligence and Public Engagement fit together?
I’m a Sci-Fi geek at heart so when I was given the opportunity to explore the benefits of Artificial Intelligence (AI) I didn’t even hesitate, even if I don’t have any of the expertise required to implement it myself such as data science, engineering, or mathematics. If you look at subjects at the top of my Sci-Fi list you’d find time travel followed by AI.
I followed progress in the world of AI with wide eyed wonder as AlphGo accomplished what many experts predicted we were still a decade away from accomplishing — a machine capable of besting human players at the ancient game of Go. This is not an easy accomplishment as there are as many possible moves in a game of go as there are atoms in the universe, so in short you can’t brute force your way to winning. Grandmasters in the game often talk about intuition when it comes to making certain winning moves, it just felt right is a common statement. So for a machine to develop this level of expertise in such a short amount of time in a game that has been around and played by humans for over 2 millennia is nothing short of astounding.
As I dipped my toes in the very cold water of AI in the GC I quickly found a strong combination — AI and public engagement. Public engagement struck me as an area that can benefit tremendously from leveraging AI, especially at a time where engagement is difficult to do well. Truth be told much of the struggle on the engagement front has a lot to do with art rather than science but that’s a separate conversation altogether. Human factors aside AI paired with a thoughtful and meaningful engagement strategy could really stand to catapult these efforts to astronomic proportions.
For example Natural Language Processing (NLP), a form of AI that processes and makes sense of unstructured text (think Watson on Jeopardy) and automated machine translation (think Google Translate) can be used to scale engagements beyond the physical realm and support large participation rates as well as allow a great many people to communicate with each other in real-time even if they do not share a common language. Imagine a future where Canadians are free to engage with their government in their native tongue regardless of what it might be. This is only possible through the power of AI, you couldn’t manually support this.
NLP paired with classification based AIs can be used to semantically enrich all citizen submissions so that conversations can be segmented into themes or groupings both so that participants benefit from a deep and evolving discussion that has many dimensions as well as benefitting analysts that no longer need to perform the repetitive and sometimes exhausting exercise of making sense of all the data.
AI can also be used to detect fraudulent submissions as well as play an active and proactive role in monitoring the tone and intent of messages so that rules of engagement are respected such as no bullying or troll like behaviour occurs further reducing the burden on our limited resources to host online conversations.
Finally, it can provide the means to make sense of huge volumes of data in real-time opening up the possibility of providing new ways to tap into our citizens’ consciousness including ways that tap into the silent majority. Obviously all of this could be further explored in greater detail and over the last year a small team at Canadian Heritage has been hard at work to make steady progress on some of these fronts. We’ve been using a platform called Pol.is to conduct online engagements which provides us a unique way to engage with Canadians while leveraging new and advanced digital capabilities.
