Aldo Agostinelli

Will technology ever replace human intelligence? Facebook’s latest mishap suggests not. The social giant’s attempt to automate their trending news section backfired spectacularly. 

Until recently, Facebook’s trending section, which shows which news stories are most popular and relevant around the world, was manually curated by a team of editors. Fuelled by allegations of political bias in the stories selected, and Facebook’s constant strive towards automation through data, the platform swapped human intelligence for an algorithm.

Just three days in, it all went downhill. A false story slating a Fox News presenter by a right-wing propaganda site slipped through Facebook’s algorithm and received thousands of shares by users before the mistake was noticed, causing outrage and controversy across Facebook. Human editors may not be able to process data or spot patterns as quickly as machines – but they can easily recognise false stories, subtleties and satire in a way that no machine is capable of.

Machine errors like this suggests there’ll always be a need for a human touch in technology. But are there any cases of computer intelligence beating humans hands down?

In some instances, the answer seems to be heavily leaning towards yes. Take driverless vehicles for example. Uber has started using them as part of their service and in Australia, on-road trials have begun for driverless buses. Both have run flawlessly so far, making their businesses safer and more efficient in the process. (Human) driver error is believed to be behind 90% of all crashes so the potential for driverless cars to reduce road mortality is incredibly exciting.

But again, driverless vehicles have had their problems, like when Google’s self-driving car crashed into the side of a bus. Interestingly, Google blame the bus driver for the crash because the Google car made an assumption about how the (human) bus driver should have been driving. On the one hand, you could argue that the crash wouldn’t have happened if all cars were driven by machines that drove exactly as expected. On the other, many would argue that any human would have been capable of the subtle negotiation between drivers needed to decide who was to go first at that particular junction. Who knows…

When it comes to human intelligence versus technology, it’s impossible to say that one will come out top. We need to make the most of both instead of going to one extreme. Vitally, we need to keep innovating. The next step is finding the best way to enhance performance through a combination of technology and human intelligence.

In some industries, technology may take over tasks that are very simple – or those that too complex for humans. For example, Facebook are manufacturing AI systems programmed to build AI because the highly trained engineers capable of doing so are few and far between.  However, there will always be a place for humans, even if it’s just monitoring automated processes. If human editors and an algorithm work together on Facebook’s trending news, we’ll get the best of both worlds.

Which side do you stand on? Will technology eventually overtake human intelligence? Which industries will it impact first? Tweet me @AgostinelliAldo and let me know your thoughts.

This blog is also available in Italian.


Aldo Agostinelli