Science & Technology

The rise of the machines

Artificial Intelligence is riding on both hope and fear. Here's why

By Rakesh Kalshian
Published: Saturday 30 April 2016

Illustration: Tarique Aziz

Last month a computer program called AlphaGo took on one of the greatest players of the complex board game, Go, and beat him four games to one. Lee Sedol, the South Korean Go master, lost $1 million, offered as prize money by Google, the challenger’s proprietor. But, more momentously, his humbling has taken the battle of wits between machines and humans to yet another level.

To be sure, this isn’t the first time software has got the better of an exceptionally clever human. In 1997, IBM’s Deep Blue had checkmated Garry Kasparov, the erstwhile world chess champion. Deep Blue triumphed because its brute computing power enabled it to dive deep into chess archives and quickly come up with the most winnable move. AlphaGo too exploits the same processing agility and thick memory but, unlike Deep Blue, it is also groomed to be an autodidact—it can become smarter by playing against itself and, just like we do, commit that learning to memory, thus creating a primitive analogue of human experience or intuition.

Many have dubbed AlphaGo’s victory as spectacularly precocious and pathbreaking. They believe that algorithms like AlphaGo will eventually make human tasks and skills that entail trawling through and parsing reams of data, such as diagnosing disease, predicting weather or even detecting art forgeries, easier and more accurate.

However, some observers, though in a minority, are not euphoric about artificial intelligence’s (AI) latest poster boy. They contend computers were in any case designed to excel in solving complex puzzles like chess and Go, and as they become more powerful and their algorithms cannier, machines will eventually make humans passé in data-heavy tasks. But luckily for us, a majority of human endeavours require unprogrammable traits as empathy, ethics, esprit de corps, and irrationality.

American neurologist Robert Burton believes it is unhelpful to pit AI against ours. Instead, he proposes that we should humbly concede that AlphaGo is a new kind of intellect, though utterly devoid of moral or emotional intelligence. So, for problems that require sorting out and interpreting gobs of data, such as managing peak-hour traffic, or making sense of census data, smart algorithms should be our go-to thing. But when it comes to making policies or laws on complex subjects like nuclear power or juvenile crimes, it is best to leave AI out of the picture.

At the other extreme, some Cassandras fear that AI could potentially turn into a Frankenstein nightmare. Critics say it’s merely a trope exploited by sci-fi writers, but there are valid anxieties over AI’s military applications. Even physicist Stephen Hawking believes that AI could spell doom for human existence. In cue with such concerns, some people have started creating moats against such Trojan horses. DeepMind, Google’s AI arm and developer of AlphaGo, for instance, has put together a safety and ethics board to ensure that such technologies are not hijacked for evil designs. All of this may sound a bit fantastically alarmist, especially fears about robots enslaving us, but make no mistake: wittingly or unwittingly, we are already caught in the ever-expanding web of AI software. What are Google search and maps, if not one of the many tentacles of this web?

The trouble is that some of this software is so efficient and smart that before we know it, we are hooked. For instance, anyone who relies on Google maps for directions while driving would be totally lost without it. That said, isn’t that generally true of our relationship with technology? It is almost always a trade-off between aesthetics and ethics on the one hand, and economics and efficiency on the other.

Besides, we are too intellectually arrogant to brook any challenge from even other species, let alone machines, which, ironically, are products of the same hubris—Kasparov blamed IBM of foul play while Sedol vowed never to play against a machine. Burton suggests a graceful way out of this Catch-22: “Rather than fretting over what sources of pride machines will take from us, we should focus on those areas where man alone can make a difference.”

Subscribe to Weekly Newsletter :
Related Stories

Comments are moderated and will be published only after the site moderator’s approval. Please use a genuine email ID and provide your name. Selected comments may also be used in the ‘Letters’ section of the Down To Earth print edition.