I hadn’t heard the term “anticipatory computing” until a little more than a week ago at Journalism Interactive.
I certainly knew the concept: using technology to anticipate human needs and then to provide ways to meet those needs. Science fiction abounds with artificial intelligence that does just that. Think of Rosie, the robotic maid from “The Jetsons,” the 1960s cartoon, who used to anticipate the needs of George, Jane and their family, with humorous results.
Anticipatory computing today is serious business, though. Amy Webb, whose digital strategy agency helps companies look two to seven years in the future, brought up the concept at Journalism Interactive, as did Jeremy Bowers of NPR. They pointed to software and apps like Google Now, Glympse and Emu as part of a trend that journalists need to pay attention to – and take advantage of.
For instance, Webb said Google Now beeped her phone that morning after it identified heavy traffic in the area where she would be traveling in the Washington suburbs. Emu uses its text messaging service to anticipate users’ interests, offering suggestions for things like movies and restaurants. It will also add events to your calendar automatically based on those texts. Journalists need to start thinking in terms of these contextual applications for reporting and for producing content, she said.
The rise of robo-journalism
As part of discussions about the growing role of computers, algorithms and robots in journalism, Webb and Bowers both brought up the 4th Down Bot, which The New York Times provided during the last NFL season. Based on 10 years’ worth of game data, it analyzed whether NFL teams should try to make a first down or punt on fourth down. It worked well with Twitter because it helped generate audience conversation on a back channel during games.
Webb and Bowers also brought up an earthquake story written by a bot that was monitoring seismic activity for the Los Angeles Times. Ken Schwencke, a Times reporter, created an algorithm that drew on data from the National Geological Survey and fashioned a story in a matter of seconds.
Webb said journalists needed to embrace that idea. Rather than fearing that computers will take over their jobs, they need to put those computers to use to do the grunt work. That allows journalists to focus on the more difficult analytic information – the what does this mean? aspects – that audiences yearn for. Other examples:
- WikiSeer, an add-on for Firefox and Chrome, summarizes complex stories, pulling out key information into a few relevant paragraphs.
- Quill by NarrativeScience looks through complicated data sets and pulls out relevant details quickly. She says it’s a great tool for helping journalists prepare for interviews.
- WolframAlpha (pro version) provides quick contextual information. For instance, she searched for “flights overhead” and got a list of all flights that were over the area at the time, along with a map. A search for “Hamlet” gave such statistics as the number of words, most frequent words, characters, dialogues and acts. WolframAlpha allows uploading of data, and she suggested using it to analyze transcripts of hearings and meetings.
MindMeld, which she said the most exciting digital tool she had seen in the last year, is like Google’s listening to what you say and spitting out answers before you can ask questions, she said. She urged journalists to have it at their desks and make it a part of their routine.
For instance, imagine doing a phone interview and saying, “I didn’t realize that David Tennant was a candidate for prime minister.” MindMeld would record your words and post them in a box on the screen. You touch the box and MindMeld immediately pulls up articles about David Tennant and prime ministers.
I’ve been testing MindMeld over the last few days and have found it amazingly accurate, even without a headset (which the company recommends using). The app records only when you turn on the microphone or say, “OK, MindMeld.” It uses the iPad location tracker to further enhance its focus, often bringing up maps and photos of nearby places.
I see immense potential not just for journalists but for anyone who is willing to talk to an iPad. (I use a lot of voice notes already, to my wife’s chagrin, so I’m hoping MindMeld won’t be banned from my house.)
A Bluetooth informational valet
Estimote Beacon is another digital device I plan to experiment with. It’s a low-power, programmable Bluetooth device that sends signals to nearby smartphones. The beacons are being used in retail stores to send messages to customers about various products.
For instance, the beacons will send signals to customers and tell them of discounts on products, show them different styles or colors, or provide options for buying. I’ve ordered some for an app project aimed at reaching out to prospective students, and I see potential in museums, libraries and anywhere else that needs to provide information to people.
I’m still getting used to the idea of computers anticipating what I need or what I might do. Using these devices and apps means giving up a certain amount of privacy, and I’m not always comfortable with that. And having a computer “know” me still seems creepy at times.
Digital first, yes, but now technology first
I long ago realized that resisting technology is futile, though. Webb even raised the idea of technology-first journalism.
The idea of digital first should be a given for journalists, she said, but digital first describes only workflow. If journalists don’t start developing technology that works for them, they will perpetually follow rather than lead the information business. That’s because journalists have no control over the people who are making new technology products or how people are using those products. News organizations must evolve into tech organizations or start to acquire tech companies, she argued.
To help explain this idea, she pointed to newspapers at the turn of the 20th century. By owning printing presses and controlling newsprint companies, they controlled the flow of information. By working closely with companies like Linotype, they shaped key technological components of information production.
She offered another example: Search bars on news websites generally produce poor results because news organizations have paid little attention to them. As a result, even if people are hooked on a story on a particular news website, they will go elsewhere to search for more information.
Companies that are using a tech first strategy today include Vox, Quartz, Circa, Huffington Post, ProPublica and to some extent the New York Times (though she said the Times’ software projects took a lot of time to trickle down into use).
If all this makes you feel overwhelmed, you are in good company. Technology is changing faster than any of us can keep up with. Just remember that you don’t have to become an expert at all of it. As I tell my students, the best approach is to keep an open mind while pushing yourself to experiment.
If you think about it, journalists have always had to anticipate the needs and wants of audiences. In this new age of anticpatory computing, they are just using technology to help.