Amy Webb is the founder and chief executive of Webbmedia Group, a digital strategy firm that advises an international client base on near-future emerging technologies and digital media trends. She is also a visiting Neiman fellow at Harvard University.
One of the best Twitter accounts — inside the Beltway or out — belongs to former representative John Dingell (D-Mich.), who announced his retirement with self-effacing posts such as “Added the ‘F’ word to my Twitter bio” and “Also retiring: the underscore in my username.” He’s also tweeted an inside joke to master Photoshopper @darth, posted incredible #TBT selfies and called out the Environmental Protection Agency for tweeting about Kim Kardashian.
Dingell gets social media in a way that has eluded many members of Congress, who don’t always stay current on technology. During the last session, they tweeted in the third person, alphabetized their floppy disks and sent thousands of e-mails from their BlackBerrys.
And yet during that time, we saw an unprecedented number of data breaches and hacking attacks in both the public and private sectors. Bitcoin, a new cryptocurrency and the system it runs on, were hardly discussed. A national debate about the future of the Internet and how we use it was ignored by too many members of Congress to list, but I don’t blame them. Net neutrality is a heady, complex issue.
It’s now or never. Technology — and the various ways we use and exploit it — is inextricably tied to governing and being governed. So congressional leaders and their staff members must embrace not just what’s on the horizon but everyday technology found in the private sector. The longer members of Congress wait to adapt to and adopt new technologies, the harder it will be for them to catch up.
I regularly meet with researchers, academics, hackers and makers to learn about their latest projects. I also review microeconomic trends, changes in consumer behavior, patent filings, new software and gadgets from around the world. Then, each December, I make a list of trends that show where technology is moving. Of the 55 big ideas highlighted for 2015, five will affect our elected leaders. Here’s what Congress should do with them:
Cognitive computing systems use artificial intelligence and natural language processing, which extracts meaning from our everyday speech, to understand human intentions. In short, researchers have taught computers to think, and they’re starting to make inferences and decisions. For example: Toward the end of 2014, Google researchers unveiled a project that used such a system to caption a bunch of photos — without explicitly teaching the system to recognize what was in them. A photo with the caption “Two pizzas sitting on top of a stove top oven” showed exactly that, plus a nearly drained glass of wine. To craft that caption, millions of digital images were processed through thousands of computers so that Google’s neural network could teach itself to recognize what was being depicted.
When IBM put its supercomputer, Watson, on “Jeopardy” in 2011, it made for great television. But Watson is more than a game-show stunt — it’s an extremely powerful platform capable of learning and thinking on its own. It can take a big heap of data and help a doctor make a difficult diagnosis. Cognitive systems can compose music and new languages, without any help from humans.
Could a computer invent a new language to help Congress have better dialogue? Probably not. But here’s what a cognitive system could do for Sen. Joni Ernst (R-Iowa), who defeated Rep. Bruce Braley in a very tight race: Her staff could use machine learning to better understand her constituents, especially the difficult ones who can list many things they dislike but cannot articulate what they want.
Robots sometimes look like humans, but often they’re just smart programs running on the Internet. A series of bots, many experimental, are being programmed to do specific tasks, such as communicate with people or make purchases. Soon, intelligent bots will be running . . . from the law.
For example: The Random Darknet Shopper is an art project and automated shopping bot. It’s programmed to spend $100 in bitcoin every week on a random purchase within a specific online marketplace. There’s just one problem: Recently, it bought 10 ecstasy pills and a falsified Hungarian passport. Last year, Snapchat had a porn bot problem that seemed unstoppable. It would indiscriminately send photos of naked women to random phones — spamming chief executives and teenagers alike. (Snapchat has prevented those accounts from continuing to post, however this kind of spam is an ongoing problem for many networks.)
Here’s what that could mean for Rep. Larry Bucshon (R-Ind.), chairman of the House Science Committee’s subcommittee on research and technology: The next wave of cybercrime may be unintentional. How can Buschon’s panel create policies for bad bots without implicating their well-meaning programmers?
As we continue to upload personal data to social networks, photo- and video-sharing sites, wearable devices, and elsewhere, companies will continue to question who owns the rights to our data — and who has the right to look at it. Ultimately, Congress will have to decide.
For example: The data collected by wearable devices isn’t necessarily private. Check the terms of service for that watch or smart pedometer you got for Christmas, and you’ll discover that the personal health information your device is collecting isn’t entirely yours. Last year, Sen. Chuck Schumer (D-N.Y.) called on the Federal Trade Commission to prevent companies like Fitbit from selling users’ data to advertisers. While the FTC considers that request, it is also looking into how Apple will shore up encryption and data protection when its watch debuts this year, with Commissioner Julie Brill saying she’s concerned about health data leaks.
In the meantime, the Blackphone, a $629 smartphone that helps block the National Security Agency, the Department of Homeland Security and just about everyone else from eavesdropping, will roll out to the general public in 2015. Last year, Sen. Bernie Sanders (I-Vt.) sent a scathing letter to the NSA’s director, questioning whether the agency had been spying on members of Congress. Maybe he should get on the Blackphone’s waiting list. Congressional leaders, such as Sen. Rand Paul (R-Ky.) and Majority Leader Mitch McConnell (R-Ky.), often argue for more digital privacy but don’t have enough information about what’s happening with our data and why.
An algorithm is a set of rules or processes that are followed to solve a problem. They’ve been around for centuries in one form or another, but now that computers are so powerful, algorithms need supervision. Algorithms promise power and efficiency, but they can also cause significant damage when left unchecked. Depending on the decisions programmers make and the data they use, algorithms can produce wildly inaccurate results. As a result, we are increasingly misclassifying objects, data and people.
For example: There are lots of stories of algorithms wrongly identifying terrorism suspects at airports, and algorithms for high-frequency trading nearly destroyed the stock market. Sens. Ron Johnson (R-Wis.) and Tammy Baldwin (D-Wis.) should ask the Senate Committee on Homeland Security and Governmental Affairs, which they lead, to invite experts from the private sector to help educate lawmakers about algorithms. This isn’t about governing or regulating algorithms, but instead not relying on their answers with complete confidence.
In the meantime, lots of our elected officials are already gearing up for the 2016 campaign season, and I predict a lot of misclassification of voters and donors alike. Algorithms can be used to pinpoint specific kinds of constituents and to predict their likely votes, but only if the data is good.
Sunday-morning talk shows are still the prize, but members of Congress are missing a big opportunity in an old form of publishing. Newsletters, podcasts and niche networks that captivate smaller audiences will make a comeback in 2015, and these one-to-few platforms are the perfect way to harness constituents’ attention. But keep in mind: The reason this next generation of newsletters and podcasts is succeeding is because they use a different format. They’re personalized and intended for a particular group of people, and therefore make readers feel more exclusive and special.
For example: Some of the newest social networks are going in the opposite direction of Facebook and Twitter. Rather than letting users post all day long to the largest audience possible, they mandate quality and editing. This.cm is one such network; each member of the community gets to post only one link a day. You can’t just join, though — you have to be invited by another member. It’s quickly become one of the hottest invites around.
Dozens of newsletters such as Today in Tabs and MediaREDEF have been backed by investors, even though they have small audiences. Podcasts are on the rise again, too. It may seem counterintuitive to pollsters, but it’s already clear that a smaller audience can be more valuable. Voters crave access to exclusive digital networks. When a critical mass of users is actively engaged in a network, that means they’re paying attention, and that’s the metric that will matter most in the next election cycle.