Status of Women Committee on Dec. 5th, 2016

[Dr. Sandra Robinson]

I don't think so, because we've seen regulation fail us, I think, in many ways when we try to contain or constrain information flows. They have a way of slipping out of the leaky boundaries we put around communication. I think the education angle is a much more powerful one.

Actually, legislating private organizations and private companies is much trickier to do, given that we are talking about many different kinds of organizations and the many different ways that they approach collating and curating information. Each algorithm is proprietary, so it's a trade secret. It's like Kentucky Fried Chicken: there's no way they're letting that out into the open.

[Voices]

Oh, oh!

[Dr. Sandra Robinson]

You might want to answer that, Corinne.

[Corinne Charette]

The other thing I would add to that is the notion of a free and open Internet. As for the notion of censoring, it's very difficult in terms of what can be done when you don't know who the user is. I'm not sure that's the best route to the end goal.

[Rachael Thomas, Conservative, Lethbridge, AB]

Thank you.

For both of you, once again, I'd like your thoughts on this question. How could we go about changing the public narrative, then, in order to make online users more aware of the impact that algorithms have on their day-to-day lives? How do we do that? How do we change that public narrative?

[Corinne Charette]

I think we do that by explaining what happens online to everyday and all Canadians, who are not necessarily all in the technology sector and therefore are less aware of what a search is really doing and of the ins and outs of different algorithms or capabilities. I think we need to educate everyday Canadians, teachers, parents, high school students, and even grade school students.

I would start in grade school with the consequences of what happens when you use a browser and what you do online, putting it in “everyday speak”, such as asking them if they would leave an unbecoming photo of themselves in the middle of their classroom. Probably not, but then why they would post it online, you'd ask, because that's even bigger than a classroom, and so on.

As a society, I think we have a big awareness challenge to take an increasingly complex digital world and to help make those concepts “real life” to people, with real-life examples for children, teenagers, and adults alike about what they're doing and what the potential impacts are going forward. I don't think we spend at enough time doing that, not at all.

[Dr. Sandra Robinson]

I would agree. I think the U.S. election actually did escalate the conversation with the public narrative and discourse around the implications of online speech, hate speech, complex arguments around what people are saying and the implications of that, and then the impact of what people are thinking when they go online. The other side of that is that it tends to show also how hard it is to regulate that speech, right?

This is a very tricky one. I think education is key, but also key is making a link between the grossly oversimplified mechanism for uploading, posting content, and hitting the heart button, etc. There's a huge distance here between that and realizing what the material consequence of it is. Each one of those iterations drives popularity. Every single “like”, “dislike”, or smiley face pushes the ranking of a story. No matter how ugly that story is or how beautiful it is, up it goes.

[The Chair, Conservative, Marilyn Gladu]

Thank you.

We're going to Ms. Malcolmson for seven minutes.

[Sheila Malcolmson, NDP, Nanaimo—Ladysmith, BC]

Thank you, Chair.

Thank you to both witnesses.

Do you have any inspiring examples for us of where algorithms were used for good to identify, call out, or to actually prevent and address cyber-violence?

[Dr. Sandra Robinson]

Sadly, no. Nothing leaps out at me. I think there's potential there for an increase in natural language processing and machine learning to actually help us identify some of the speech that goes on in the public Twittersphere and on Facebook. That might over time improve our ability to catch these sorts of hateful events as they happen, but it's an extremely difficult thing to do. Algorithms are only as smart as as the people who train them and the data that's fed into them. It's pretty tricky.

[Corinne Charette]

Yes.

[Sheila Malcolmson, NDP, Nanaimo—Ladysmith, BC]

Is that a “no” from you also?

[Corinne Charette]

I know that there are algorithms online that present a lot of good, but not necessarily in combatting cybercrime. For instance, in the field of medical technology, they are doing a lot of big data work and even finding specific individualized potential cures and so on. There are a lot of great applications of big data in a positive sense, but I can't think of any of them in a cybercrime concept.

[Sheila Malcolmson, NDP, Nanaimo—Ladysmith, BC]

In looking for more good news in this conversation, have you seen some good examples of tools where the Internet community itself, using the human approach, identifies, calls out, and combats cyber-violence?

[Dr. Sandra Robinson]

Yes. There are some key examples of it that have occurred on Facebook. Groups of women did come together to organize a response to Facebook's refusal to remove gender-based violent images and content. It took quite some time. It was started by a group in the United Kingdom. They were very successful at pushing back on Facebook.

I think we've seen some of this happen more recently in the last four months on Twitter as well. I think an organized speaking-back campaign can make a big difference. It has to marshal not just individuals but often advocacy groups, which can get a little more attention because they can cross over many different kinds of social media platforms.

There are some stories out there, even those that don't directly intersect with cyber-harassment and cyber-bullying, such as women who spoke back against Instagram and Facebook for taking down their family photos of themselves nursing, but women who are nursing are being picked up by algorithms that are searching for nudity.

There are ways in which people do organize to push back on violence online and on violent speech and hate speech. They have been successful.

[Sheila Malcolmson, NDP, Nanaimo—Ladysmith, BC]

How do you separate those actions from feeding the trolls, which we are not supposed to do?

[Dr. Sandra Robinson]

One of the things they did, for example, was that they didn't speak back strictly by using the same hashtags in campaigns. They went directly to Facebook executives. They went directly to the leadership of Twitter. They went directly to the leadership at Instagram. I think that's really what we need to do in terms of being organized: to speak back to those who have the power to change the way information is organized on their particular social media.

[Sheila Malcolmson, NDP, Nanaimo—Ladysmith, BC]

So far as our final report goes, which is coming up pretty soon, do you have any recommendations that would make your heart sing if you read them in our final report? What would you be delighted to see as our recommendations to the federal government?

[Dr. Sandra Robinson]

Personally, I'd love to see a commitment to advancing this issue through education, right from primary school through to high school and well beyond. I use MediaSmarts materials in my courses from the first year of university through to the fourth year, and they're outstanding. They reach students. They speak to students. Students go after those materials themselves. I'm an educator, and I'm a firm believer that this would be appropriate.

[Corinne Charette]

I would certainly agree with Sandra. I think whatever the federal government can do to sponsor, champion, and encourage education with young girls and young boys, starting in grade school, in methods, a vocabulary and a vernacular that they can relate to, is so important, including for parents. I think the opaqueness of this issue is that it's so easy to click online and not understand the ramifications of any click, selection, or search that you put in.

People need to know this as much as they need to know how to drive or do banking. I think we have to put a big emphasis on this in our society. As computing capabilities continue to grow and as data continues to grow, the awareness of the users of this technology needs to keep pace with it. We need to become much more sophisticated so that we can be critical about what we do online, how we do it, what tools we use, and what tools we don't use.

[Sheila Malcolmson, NDP, Nanaimo—Ladysmith, BC]

Thanks, Chair.

[The Chair, Conservative, Marilyn Gladu]

We're going now to Mr. Fraser for seven minutes.

[Sean Fraser, Liberal, Central Nova, NS]

Thank you, Madam Chair.

Before I start with the questions, I would like to thank the witnesses for being here today.

I have a question for Ms. Charette.

You spoke a bit about the innovative practices and all the great applications that exist, which seemingly nothing to do with violence against women specifically. I'm very excited about the further development in Canada of a personalized user-based web experience. I think there are a lot of great applications, both from the consumer's perspective and from a business perspective.

Of course, I share the same concern the committee has, which is that the Internet seems to have gotten very good at showing us the good things that we want, but also at showing us bad things, whether we want to see them or not.

As we're seeking to make recommendations to the government, I've heard you loud and clear: education has to be at the fore when we're making these recommendations. Are there certain areas that you'd like to see us avoid to make sure we don't stifle the positive innovative practices that are developing in the private sector?

[Corinne Charette]

I think that in any dialogue it's important to cover all sides of the story, so we wouldn't want to fall into strictly negative dialogue on big data, big data analytics, and the negative potential. I think we need to keep stressing all the good things about big data, including predicting when there are weather events happening, and also in terms of how, based on social media, communities can come together quickly in an emergency to solve a problem and to help find missing people and so on.

I think we need a balanced dialogue that highlights the great examples of the positive use of big data analytics and the great innovation potential for our economy from firms that are developing these tools, collecting big data, and making open data available. The federal government is a big sponsor of open data, through the open data portal. We have many open datasets, a lot of which are geospatial, which help Canadians, as well as many people internationally, understand the different constructs of the country,

[Sean Fraser, Liberal, Central Nova, NS]

That's excellent.

Ms. Robinson, do you have anything to add? I want to make sure that we cover off the potential pitfalls here if we're going down this path.