Why We Need to Get Smart about Artificial Intelligence
We’ve long been fascinated by the concept of artificial intelligence (AI), but only recently has technology advanced sufficiently to bring it into our daily lives. There’s Apple’s Siri and Microsoft’s Cortana; Tesla’s driverless car that obeys speed limits and parks itself in your garage, and IBM’s Watson, helping cancer doctors use genomic data to personalize patients’ treatment plans. On the surface, it sounds cool – and valuable in its potential to identify risks and anticipate problems. But as interest, investment and experimentation in the area explodes, some of the world’s leading thinkers and entrepreneurs are expressing concern about the dangers of “machine thinking” and the future impact of super intelligence on humanity.
Harvard Law Professor and renowned cyber security expert Jonathan Zittrain doesn’t share their angst. As he recently articulated in a special lecture titled “Love the Processor, Hate the Process,” AI isn’t a threat – but it does present a puzzling set of new challenges that need to be figured out, not just within law but across multiple disciplines.
Technology is evolving rapidly, pushing the line between tool and friend, explains Zittrain, director of Harvard’s Berkman Center for Internet & Society, the research institution he helped found 18 years ago. Innovative systems and algorithms allow computers to learn about us, and use sensitive data to sway our decisions and even influence our biases.
Take Facebook and Google, which utilize increasingly sophisticated algorithms that contribute to what Facebook presents in a feed, or what Google shows us in search results. Last summer, Facebook’s algorithm controversially (if unintentionally) amplified the (in)famous ALS Ice Bucket Challenge compared to, say, news of unrest in Ferguson, Missouri. Yet, it’s likely that some of us wouldn’t object to Google changing its algorithm to omit extremist websites in response to searches looking for exactly that (though Google has traditionally refused on principle to do so).
Zittrain calls many algorithms “roulette wheels” – so complicated and powerful that even the designer/operator can’t anticipate outcomes or consequences. One easy fix, he argues, is to facilitate competition. Another idea, he says, is for web companies entrusted with personal data and preferences to act as “information fiduciaries,” similar to doctor/patient or lawyer/client relationships.
The bottom line is this: The world of data and algorithms is complex, and it’s shifting constantly. We can aspire to technology’s computational fruits – if we remain aware of their implications and minimize the risks.
Zittrain has long studied the legal, technological and world-shaking aspects of quickly morphing technology terrains, and continuously seeks opportunities for a better world. One of Zittrain’s renowned ideas – Perma.cc, a Harvard Library Innovation Lab project that takes on the problem of “link rot” or broken or defunct links in scholarship – was recently honored with a Webby Award for best law site of 2015. Recently named the George Bemis Professor of International Law at Harvard Law School and the Harvard Kennedy School of Government, he is also a professor of computer science at the Harvard School of Engineering and Applied Sciences.