From The Desk of the Media Strategist: Focusing on the Bigger Story


For all the talk of new channels, content types and communication strategies, the heart and sole of public relations remains media relations. And while there have been many talks, books, guides, etc. on pitching media, it often comes down to knowing ahead of time what a reporter covers and the content that is most valuable to them.

But remember that pitching a reporter is only half the battle. If they’re interested in your angle, they must now sell the story idea to their editors. To help us understand how stories evolve and what will entice an editor, who better to turn to than Scott Dadich, editor in chief at prominent business and technology publication WIRED.

In a recent issue, Scott provided insight into the editorial process at WIRED and how stories are determined. Below is a transcript of Scott’s feedback to the story pitch submitted by business writer Cade Metz. It’s an interesting read as this is exactly how PR practitioners should think when constructing pitches – and an overall reminder to always think bigger picture in order to help your favorite reporter get their story “green-lit.”

As appeared in the June 2016 issue of WIRED:

Last October, Fan Hui walked into a six-story office building near King’s Cross station in London, headquarters of a Google-owned AI startup called DeepMind, to play a game. Hui is the European champion at Go, the 2,500-year-old test of strategy and intuition that makes chess look like checkers. Its black and white counters have more possible positions on the Go grid than there are atoms in the universe. Every move has more possible outcomes than even the most powerful artificial intelligence had ever been able to calculate. Human grand masters don’t play the game by looking ahead at possible future moves. The move according to how the board looks. A machine needs more than “brute force” to beat the top humans. It needs something closer to, well, human intuition.

EIC Commentary: Cade had already been covering AI and this Go story, so we knew he was well sourced and steeped in the material.

Hui was in London to meet just such a machine. Built by a team of Googlers led by an engineer named Demis Hassabis, the AI relied on a technology called deep learning, a mimic of the interconnections in a human brain called a neural net. Feed it enough photos of a cat and it can learn to identify a cat. Feet it enough spoken words and it can learn to recognize the commands you bark into your phone. Feed it millions of Go moves and a neural net can learn to play Go.

EIC Commentary: The creation of a new kind of artificial intelligence I loved that. I, for one, welcome our new robot, etc.

Now, in theory, that would only produce an AI that’s as good as human – not better. So the team went a step further. They matched this neural network against itself. Two (slightly different) versions of the system played each other thousands of times over, carefully tracking which moves took the most territory on the board.

EIC Commentary: We’re always looking for inflection points, places where the future is starting to happen. That’s what this pitch argues.

The system that resulted, called AlphaG, is what Hui was there to play. In view of a few members of the DeepMind team, an editor from an academic journal, and an arbiter from the British Go Association, AlphaGo and Fan Hui played five games of Go that week. And AlphaGo won them all. In March, AlphaGo is going to South Korea to play Lee Sedol, the top Go player in the world. Experts – in Go and AI – think it’ll win.

EIC Commentary: Here’s where it goes from a topic to a story. Now there’s a conflict in the narrative and a moment we knew Cade could focus his reporting on: a human genius up against a Cylon, with the future of humanity in the balance.

The game though, is really just a proxy war. Deep learning has already proven adept at identifying images, recognizing spoken words, and even understanding natural language. AlphaGo’s abilities point the way to a future where robots interact with the physical world the same way the system interacts with Go – learning from its environment and responding to unexpected changes. As DeepMind built AlphaGo, Mark Zuckerberg and his AI researchers at Facebook were using deep learning to build their own Go player. This fight is really between Google and Facebook, over who will build the first intelligent, adaptable computer.

EIC Commentary: And these are the stakes. Beyond the metaphysical, building these AIs will have implications for the entire tech industry.

I’ve been covering this story as it unfolded, and I have exclusive access to the DeepMind team in the run-up to the match with Sedol and during the match itself. I propose building a story that spans not only the path of AlphaGo, from inception 19 months ago to the Fan Hui match to the match with Lee Sedol, but also the recent history of AI – a field that is moving faster than anyone, even its most prominent practitioners, expected. Structurally, I’d set it around a series of Go matches: AlphaGo versus Fan Hui. AlphaGo versus Lee Sedol. AlphaGo versus me. And, if I can set it up, AlphaGo versus whatever they’re building at Facebook.

EIC Commentary: (Part I) Bricklaying for the story. This spells out for the rest of us how Cade and Marcus plan to tell the tale. (Part II) A great idea that we didn’t use in the end. Cade’s access to the tournament in South Korea was too good. You’ll see.