I was at the South by Southwest (SXSW) event earlier this month because it is the best event for getting a sneak peek on what will come. Future-oriented technology leaders and doers gather in Austin, Texas, to share their views on the future and what they are currently doing to shape it. Yes, it is also a film and music festival, so it also attracts a lot of creative thinkers.
From what I saw, this year wasn’t really about how tech is developing, but rather what the tech development impact is on society and humanity. The big question is how to make sure the future is not self-driven but controlled.
The most used example was, in fact, fake news. It demonstrates a point about platforms and tech developers being responsible about what their impact on society is. Suddenly legacy news media brands are now valued for editorial responsibility and cool in their trustworthiness (something considered a bit boring earlier), and tech platforms are evil tech, not taking enough responsibility.
One of the hot technology topics is, of course, machine learning and its applications. Machine learning was said to be the new mobile — it defines what is being developed.
But also related to machine learning is the key topic of responsibility. Machines learn to be biased when they are given biased data. Google’s image recognition marked black people as gorillas, a “racist” soap dispenser failed to give soap to a black hand, and speech recognition generally does a much better job with men’s voices than women’s, even though women tend to speak more clearly than men, on average.
All of this is estimated to be due to bad teaching data.
So, the key takeaway from Artificial Intelligence (AI) design was to be responsible about the teaching data. Whatever bias there is in the data that goes in will come out biased as well. And when this happens, be prepared for a PR disaster.
And why settle for AI that works just like the world currently behaves when it can be trained to work like the world should be?
Relevancy of this discussion for news media is reflected in our AI-based recommendations:
- Does our teaching data for recommendations include enough information about different kinds of people?
- Does the recommendation attempt to give a reader more of the same that the reader has read before? Or does it make sure that reader sees the opposite perspectives too?
With news personalisation, we can create bubbles or break bubbles, depending on the design.
In the marketing stream, the most talked about topic was purpose and societal impact of brands. Now that societal responsibility is what consumers are also interested in, purpose-driven brands are said to be prospering. There is also a more critical eye on being good only on the surface. Conducting charity campaigns and initiatives may seem like whitewashing compared to those companies that offer products doing good for society.
Those truly purpose-driven companies are doing better, especially with Millennials; 60% are said to be belief-driven buyers and 46% say they would spend more for brands that lead with their purpose.
I thought this was really interesting, as news media companies like ours are profoundly about giving people information and insights about the world so readers can make more informed decisions. This is a great reason to exist, a great purpose. In today’s fake news world, we have seen glimpses of what the world could be without quality media.
PepsiCo’s vice president of marketing, Todd Kaplan, talked about purpose-driven brands — they have their purpose tied into the reason for existence. He talked about the LIFEWTR product as a good example of a purpose-driven brand. The company had created this bottled water brand to support female artists.
Bottled water? I thought it was crazy to portray a plastic water brand as doing good for society, when we know the environment is in trouble because of plastic and water bottles are a big part of it. But, hey, if a bottled water brand markets itself as being purpose-driven, we definitely should consider doing it, too, especially if we also make sure the AI bit in it works for the good of society in our hands — not creating bubbles but bursting them.