AUSTIN, TEXAS -- I’ve been here at SXSW for two days and I must say, it’s been amazing. The best part about all of this so far, aside from the panel sessions and workshops, is walking around the city, meeting random people and hearing their stories.
The track I’m following is Content and Distribution, and the most popular debate so far (based on the sessions I managed to get to) has been about human- vs. machine-generated content. Which made me wonder: Who actually creates better content that will truly connect and resonate with audiences?
Day One – Can a machine define beauty in photography?
On day one, Ramzi Rizek, co-founder and CTO of EyeEm (a stock photography site) and Miriam Redi, a research scientist from Yahoo, went head to head in an animated debate with Danielle Strle, director of culture and trends at Tumblr, about the definition of beauty in photography.
Rizek and Redi both worked on Deep Learning, a type of machine learning based on algorithms that teaches machines to think and process information (in this case, beauty in photography) the way a human mind does.
Rizek took that and developed EyeEm Vision, a Deep Learning-powered Computer Vision framework that can accurately recognize more than 20,000 individual concepts AND rank photos based on their aesthetic qualities. That’s pretty amazing, and scary at the same time.
Because the thing is, beauty is such a subjective and fluid concept that’s constantly evolving, which is precisely what gives it the human connection. A photo that I may find aesthetically pleasing may not evoke the same emotions in the person next to me.
So then how can a machine be able to rank photos based on their aesthetic qualities and actually be accurate? If anything, this technology is best used for sorting through the ginormous amount of photo content generated daily and sorting them out into different buckets. That would be an amazing application of this use of data and analytics.
And that is exactly how Strle felt (that’s why the animated discussion). At Tumblr, she and her team of editors sort through thousands of pictures shared daily to seek out interesting content to promote. She admitted that it’s a tedious task, but it’s something she wouldn’t give up because of the debates her team goes through daily trying to each fight for their case on why their content is more engaging than the others. Defining beauty is a human thing, and it is supposed to bring out human emotions, regardless of whether it’s in the editorial team or the audience.
So after that session, I couldn’t help but find myself aligning more towards Strle rather than Rizek and Redi -- but the thing is, I also see value in what they are doing with the technology.
Day Two – Content is about connecting, and you need humans for connection!
The two sessions I attended on day two actually showed both sides of the argument about human- vs. machine-created written content. The first one I went for was about the use of big data and analytics in the world of journalism.
Executive editor Martin Baron, and Shailesh Prakash, CIO and VP of technology, at The Washington Post, talked about their new baby, Project LOXODO: a predictive analytics platform that taps into data and analytics to help them create better and more relevant content that reaches their readers in a more timely fashion.
The technology itself isn’t something new, but the application of it, in this case for a newspaper, is rather fascinating. I’ve seen examples of data and analytics being used on the marketing front to push more relevant content to users, but seeing it used for journalism puts the technology in a completely new light.
In journalism, it’s not just about getting the relevant content to readers, it’s also about getting it to them in time. News gets old fast. So especially in a time when content is being churned out online at a ridiculously insane pace, being able to sift through the noise and finding the right content that’s delivered to you on time is pretty damn amazing.
On the other hand, the second session I attended was about fighting against machines to have more human-generated content, because that is what clicks with readers. Kate Lewis, digital media VP and editorial director of Hearst magazines, Locakhart Steele, editorial director of Vox Media Inc., and Neha Gandhi, EVP of editorial strategy of Refinery29, shared their take on how, at the heart of good engaging content, is a human writer making human connections.
This is not something a machine can and should do, they said. No matter how much Deep Learning programming a machine goes through, it will never be human. And if it can’t feel the same emotions as an actual human being, how can it connect with another human?
And here, again, is where I start feeling like hey, that actually makes a lot of sense (must be the human emotions getting to me). Because as much as I feel that technology does a lot to help find more relevant content from the crazy amount of content thrown online everyday, relevant doesn’t always mean riveting. Written content that connects on a deeper human level can only work when another person creates it. Like beauty, human connections are fluid concepts that constantly change from one person to the other and because of this highly subjective nature, it’s impossible for a machine to fully grasp.
So what’s the future of content?
From what I’ve managed to see in this past two days at SXSW, I’ve realized that as much as technology is advancing, you still need a human touch to create content that’s engaging and relevant to readers. In fact, at the rate that technology is growing, we might actually need more humanness (is that actually a word?) to make sure that our content will connect on a real level with other people.
But you still can’t discount the changes that technology has made in the way we consume and distribute content. Technology has played such a big role in creating content that’s more engaging (whether it’s through interactive videos, articles or 360-degree videos) and also in the way we sort through the huge mess of stuff online to find what’s most relevant to us.
So I guess what I’m trying to say, amid all this rambling, is that technology has its place in the future of content, but we shouldn’t neglect the human aspect of creation.
As I said at the beginning, one of the most interesting things I did at SXSW was meeting strangers and listening to their stories.
Whether it’s the cashier telling me about her day, a bunch of kids pushing their new app -- or even a homeless junkie fresh out of prison trying to bum a cigarette off me, these are all human stories that made feel happy, excited and scared respectively.