Can editors and reporters crack TechCrunch’s real-time conundrum?

One issue repeatedly crept into a roundtable discussion when FriendFeed, Google and other high-profile web companies met at Techcrunch’s Real-Time Stream Crunchup last week. Beneath all the tech talk, it appears that the recurring issue was the very same thing that journalists have solved on a daily basis: parsing a deluge of raw information to make it useful for public consumption.

Information now arrives — and is instantly published — through various real-time streaming platforms like Twitter. Traditional print journalists had little to do with live streams, but social media is changing the newsroom, and this technological conundrum presents an opportunity for journalists to lead the way by applying their skills to filter the real-time web.

Definition of “real time”

When Steve Gillmor, podcaster and TechCrunchIT editor, asked panelists to answer a question he had heard during the daylong event — where is real-time going? — some of the answers resembled tasks that news organizations have traditionally handled.

Answers: Andreas Weigend, formerly of Amazon.com, says real time is about more than just moving data quickly. “it’s about quickly getting questions answered, and gathering data to support decisions.” The opportunity I see:  Reporters can work with developers to create a system to detect questions and match them up with questions, or enable reporters to quickly provide those answers. Daily reporters have spent their whole careers predicting and identifying important questions, and quickly finding answers.

Preservation: Live real-time streams are just one side of a larger concept that activity stream junkie Kevin Marks refers to as “flow.” At the other end of that flow, is the preservation of things that were once real-time, so that they can be found later. The opportunity I see: Reporters are constantly researching public and archived content. They can help design systems to store real-time presentations.

Filtration: For Tweetmeme CEO Nick Halstead, it’s “all about real-time filtering; getting the right content in front of the people who actually want it in real-time.” The opportunity I see: Editors can work with developers to create a curation system to facilitate the filtration of live streams. After all, editors have traditionally pruned stories to fit into the space or time allotted, and they have chosen what appeared on A1 or the 6 pm broadcast.

Humans needed

The idea of filtration came up so frequently that TechCrunch’s Erick Schonfeld finally turned it into a formal topic of discussion, asking of there is a “different way to sort of bubble up this information, that makes it less overwhelming but more immediate?” He wondered if the problem lied in the interface.

Bret Taylor said that FriendFeed, the company he co-founded, spends lots of time thinking about interface. He isn’t sure that anyone has created the ideal interface that “mixes that very palpable sensation of being realtime and the sense that it’s all relevant”

Brand trust: Marks explained that associating information with people we recognize — using something as simple as a photograph next to data — “lets us engage the filters in our brains to decide” what is important. As Taylor put it, “if you know who made the recommendation, it has social value to you.” The opportunity I see: Journalists create personal “brands” when they interact on social platforms. Make those trusted brands and friends and followers will associate information sent by journalists as important.

Genuine interaction: Personal relevance can also be derived by algorithms that examine the social history of a person and her friends. Taylor mentioned that these types of recommendation engines have largely failed, “mainly because they are cold and inhuman.” Taylor’s ideal model would be “something algorithmic that doesn’t lose that sense of humanness and genuine  social interaction.” The opportunity I see: Back to establishing a brand. That brand cannot be built founded on techniques derived from statistics and best practices, it has to be built on genuine interaction.

Difference of opinion: But social recommendation algorithms can’t deduce why people do the things they do. When you buy a romance novel for a friend, Amazon might start recommending similar books, even if you abhor the genre. Gillmor brought up another problem: He’s network friends with his daughter but their cinematic tastes are completely different. An algorithmic system probably wouldn’t detect that, and might start suggesting that Gillmor watch some films that his daughter wants to see. When Weigend suggested adding user-entered metadata to indicate why the message was sent, Halstead replied that people have no desire, and won’t take the time, to add that field, so “the computer can’t do the filtering.” The opportunity I see: Again, back to getting into establishing a social brand through genuine interaction. If people get to know journalists, they can identify their differences and similarities, and apply that relationship to place value on messages.

How important?

So why should journalists care? Is this just another hyped tech trend?

Expectation: Taylor said that tools like Twitter, Friendfeed and Facebook “are changing users expectations such that they expect their product to be realtime.”

Products: Taylor also noted that if a user can’t explain what they’re seeing it will be a difficult product to sell. This aligns with what Iain Dodsworth‘s experience. When he started having  problems consuming all the data being thrown at him on Twitter, he created TweetDeck. The one-year-old startup has since raised $2.3 million in funding.

Investors: George Zachary of Charles River Ventures described real-time as “fundamental to what we’ve been doing in terms of internet investing as anything in the last 15 years.” David Hornik of August Capital focused on one particular aspect: “managing and filtering the overabundance of information is what’s really going to be interesting.”

Community: Seesmic creator Loic Le Meur said that people are beginning to realize “they have communities around themselves, and it’s the first time we see them. The communities are very old, but it’s the first time I can go to twitter.com/Erickschonfeld and see who is your community.”

Steve Gillmor said that Friday’s event was just the tip of the real-time iceberg. “We are in the full throes of a gold rush that’s going on here,” he declared as the daylong event came to a close.

By the way, here is a relevant video from last year when Clay Shirky declared that we aren’t suffering from information overload. Instead, we’re victims of filter failure.

Linked to from this post:

Reach some of the people mentioned in this post on Twitter: @kevinmarks, @Erickschonfeld, @aweigend, @nickhalstead, @SteveGillmor

2 thoughts on “Can editors and reporters crack TechCrunch’s real-time conundrum?

  1. I like that term — "filter failure."

    Many people don't even know how to use basic tools — e.g. RSS readers, Tweetdeck — properly. Basic filters here can make the Web 2.0 experience much more meaningful and streamlined.

Comments are closed.