AI Researcher Alona Fyshe on the Power of the Connection Between Humans and Technology

When I started at university, my focus was fine art. But I took a computer science class as my science elective and it was the most fun homework of any class I had. I found it was both something I was good at and also something I was really interested in. So I ended up switching my degree.

And while some people might see fine art and computer science as opposites, I see many similarities. In computer science, we use creativity and problem-solving. Further, there are a lot of ways to correct problems; you can choose your own adventure and be creative and come up with different solutions—not just one. I’ve always been the kind of person who loves to make, build, and create, and those are infused in both fields.

Now I teach computer science and psychology at the University of Alberta, with research focusing on natural language processing, machine learning, and neuroscience. Here’s how I think artificial intelligence can help us all unlock our highest potential right now and into the future.

Freedom to Soar

During the industrial revolution, people were very worried machines were going to take over and nobody was going to have a job anymore. In fact, machines really did change the sort of work that people do—but I’d say it improved our work. Now, you’re able to do things that are more interesting and more cognitively appealing because we have machines that do the boring stuff.

Looking ahead, I believe AI will do the jobs that we didn’t want to do anyway. Think about the things that you have to do on a daily basis that aren’t enjoyable, or they’re repetitive and they don’t really engage you. For instance, let’s say I have to make an appointment for a haircut, and that requires me to pick up a literal phone. I might not enjoy that, and the task might be something I can’t even do if I’m at work anyway. Well, now there’s a Google program that can actually place those calls for you. It knows your calendar; it can help you navigate the ins and outs of talking to somebody and making that appointment. Then you could use that time and effort for something more important instead.

Or you’ve probably seen how Gmail will suggest a brief email reply for you. It’s a micro change that might save you a few minutes—but that adds up to a lot over the course of a day or week. Those are two very small ways in which AI has already stepped in to make our lives a little bit easier and help us use our energies more constructively.

Addressing Fears About AI

Some of the fears about AI are simply erroneous. For example, AI can’t just decide to keep itself on if you want to turn it off.

But while that type of fear is unfounded, others are more grounded in real-world applications. For instance, there are definitely applications that will write essays for students. It’s a little concerning—but people should understand that analog services for outsourcing essay writing have been around a long time, first of all.

Plus, I think it might encourage educators to think of a more creative writing process. Is there a way that writing can be assisted by AI in a way that allows us to create different or more interesting results? Sure, if it’s about the craft of essay writing, we should probably put students in a room with no technology. But if it’s something that we want to test that’s beyond the actual words on pages—higher-level thinking, structuring arguments—AI can assist with the foundation for that deeper expression.

AI can also be a threat in a way that is more insidious, in that it is built upon data that comes from humans. Therefore, it can encode the biases that are inherent—and that is something we need to be really careful about.

In this way, AI could enable the continuation of the same sort of biases and problems that our society has had for millennia. We’ve tried to move past some of these, but if you have computer programs that are continuing to promote certain biases, that can have an effect on our society.

Consider that it’s humans generating the data on which the model is trained. So if you train a language model using all of the essays that have been written in the last 100 years, AI language models would probably assume all doctors are men. These are the sorts of things that can be hard to take out of a model because of the racism, sexism, and classism in our language.

But we’ve become aware that it’s happening, and our burden is to fix it. This is a newer area of research, but it’s something that people are very interested in and concerned about—and they should be. Through these kinds of advancements, we can further develop AI technology that helps us really reach our highest ideals—and unlock our maximum human potential.

---

Canada offers business events hosted in the country a national network of thought leaders and industry champions that span industries and ecosystems. Connecting with key influencers to enhance conference programming, augmenting your speaker schedule and developing customized technical tours is as easy as connecting directly with the Destination Canada Business Events team.

(Handout photo)