Skip to content
Join our Newsletter

Machines making people look like ants

Geoff Olson Columnist [email protected] Over the years, the paving stones on my home’s walkway have sunk into the ground at off-level angles, so the other day I picked up some bags of sand to even out the dirt below them.

Geoff Olson Columnist
[email protected]
Over the years, the paving stones on my home’s walkway have sunk into the ground at off-level angles, so the other day I picked up some bags of sand to even out the dirt below them.

I hefted up the first paving stone, which turned out to be the rooftop of an ant colony.

“Damn,” I muttered, as I watched the six-legged commuters scuttle along their now-exposed highways. These weren’t carpenter ants or pests of concern to me, just a bunch of little beings going about their business, incapable of comprehending the sudden light and the vast being disrupting their world.

I slowly lifted up the other paving stones. Some of the workers were transporting larvae along the winding road system. “Sorry guys,” I said, as I emptied the first bag of sand onto the ground.

Later that day I came across a tweet from Elon Musk, chief product architect of Tesla Motors. “Hope we’re not just the biological boot loader for digital superintelligence. Unfortunately, that is increasingly probable,” he wrote.

Musk had been reading Superintelligence: Paths, Dangers, Strategies by the respected Oxford philosopher Nick Bostrom. In the author’s view, within a few generations we humans may find ourselves in the position to artificial intelligence that insects are to us. Once robots are capable of building even smarter robots, all bets are off — especially if machine consciousness results.

If the creations have obscure plans in which the creators do not fit, or decide we are a threat, the United Nations or White House might have as much ability to bargain as a kid’s ant farm on the way to the dump.

Unwarranted cyber-pessimism? Here are four items demonstrating how fast things are progressing on the AI front:

  1. Automated software programs — robots, in a sense — now account for the majority of website traffic, according to the Internet firm Incapsula. Some of it is composed of bots programmed for malicious activity. Some of it is composed of hacking tools looking for a vulnerabilities on websites. Others are scrapers, and chatbots masquerading as human beings. Only 49 per cent of website traffic is from actual people browsing the Internet.
  2. Last year, the UN and Human Rights Watch advocated a treaty banning “autonomous killing machines” such as drones capable of launching airstrikes against targets without human decision-making. Such technology is on the near-term horizon.
  3. Google has been on a spending spree, buying robotics startups and hiring AI experts to construct what some are calling the Manhattan Project of AI: basically, an all-knowing electronic Golem that makes today’s search engines look like stone tools.
  4. A neural network of 1,000 computers at Google’s X lab has “taught itself” to recognize humans and cats on the Internet, according to a report on Slate. The computers learned a “slew of concepts that have little meaning to humans.” For instance, “they became intrigued by tool-like objects oriented at 30 degrees, including spatulas and needle-nose pliers.”

Bostrom is concerned an intelligence greater than our own will start displaying behaviours that make sense only to it, but not to us. If and when machine consciousness arises, we may not even recognize it as such, even after things go sideways (or 30 degrees) for its inventors.

To say such a development would leave humanity behind the eight ball is to underestimate how badly we’d be snookered, by several orders of magnitude. We’d be the playthings of Job’s mysterious, arbitrary God, but without the anthropomorphism.

In May, Cambridge physicist Stephen Hawking penned a paper with three colleagues, noting that “it’s tempting to dismiss the notion of highly intelligent machines as mere science fiction. But this would be a mistake, and potentially our worst mistake in history.”

“One can imagine such technology outsmarting financial markets, out-inventing human researchers, out-manipulating human leaders, and developing weapons we cannot even understand. Whereas the short-term impact of AI depends on who controls it, the long-term impact depends on whether it can be controlled at all,” wrote Hawking and his coauthors.

Perhaps these academic worries are exaggerated and misplaced. But ponder for a moment how we humans treat creatures less intelligent than ourselves. Even me: I felt benign toward the ants under my paving stones, but I buried their city-complex anyway.

geoffolson.com

 

$(function() { $(".nav-social-ft").append('
  • '); });