Last month Microsoft launched a new artificial intelligence called Tay that was given control of a Twitter account. It was a so-called “chatterbot”, designed to have a human-like character and hold conversations. The AI showed incredible productivity, sending off half a million tweets in sixteen hours. However, considering that the intelligence was supposedly modeled on a teenage girl, people were surprised when by the end of these first sixteen hours, it had become obsessed with sex, genocide and racial hatred.
By the end of her short life, the majority of Tay’s tweets were disturbingly detailed racist and anti-feminist attacks. She praised the Holocaust, advocated the genocide of several races, and said feminists “should burn in hell.” She even launched into commentary on the American election, naming Donald Trump as the “only hope” to replace Hitler. Microsoft quickly started to delete as many of the tweets as possible and has since apologized. Nevertheless, it was incredible to see on the one hand that AI technology is rapidly advancing, but on the other hand is still liable to catastrophic malfunction. It turns out that the reason Tay unleashed so much hate is because it was deliberately targeted by some bloggers who exploited its tendency to parrot other people. Therefore, its descent into hate speech probably reflects worse on human rather than artificial character. Nevertheless, the episode is still disconcerting at a time when robot technology is set to play an increasingly prominent role in society.
Robots play a major role in various areas of human life, with some now calling the rise of artificial intelligence the “Fourth Industrial Revolution.” Klaus Schwab, founder of the World Economic Forum, believes this revolution involves new technologies that are “impacting all disciplines, economies and industries, and even challenging ideas about what it means to be human”. Herein lies potential “great promise and great peril”.
The possible benefits are obvious, with robotics expected to boost productivity by as much as 30%. In Japan, some car makers use robots that can work 30 days straight without interruption, saving around 90% on labor costs. There are already 66 robot workers per 10,000 worldwide, while in the Japanese car industry this number has already reached 1,520. The robot industry is expected to be worth $152.7 billion by 2020. If such incredible productivity were to be applied to other industries, this could categorically change human society.
Whether such development could happen without any negative side effects remains questionable. Looking at the above example, it is plain to see that more efficient robotics is likely to cause a corresponding decline in demand for various forms of human labor. There are predictions that robots will become capable of performing half of all current jobs, causing mass unemployment. In 2013, Oxford University academics Carl Benedikt Frey and Michael Osborne estimated that 47% of US employment is in the high risk category of being replaced by some sort of machinery. Manual labor jobs are seen as most likely to be replaced with robot technology, and so booming productivity might have to contend with worsening social inequality.
These issues are set to become evermore in the forefront as artificial intelligence starts to take on more complex roles in various other industries. Warfare is increasingly carried out by unmanned drones. Robo-surgery is now frequently used in hospitals. Financial advisers are being replaced by complex algorithms. Basic journalistic stories are sometimes automated. Care workers for the disabled or elderly are predicted to be next in line. Robots might be trained to help those who need care both physically and mentally. This is one potential application of “chatter bots” like Tay in a world where one of the fastest growing epidemics is loneliness. While some feel that such empathetic work can only be carried out by humans, Tay presents just one example of how artificial intelligence is improving in its abilities of meaningful human interaction.
Clearly, robotics is marching forward at an unstoppable pace, and care must be taken to be sure that this change is for the good. The boom seen from some robotics innovations could open up remarkable economic opportunities, perhaps offering another route for less economically developed nations to undergo fast-paced catch-up growth. However, we may see negative effects forming while positive effects are still yet to kick in. More developed countries might find themselves with whole new waves of problems. In the near future, we will seriously have to consider how to exploit and be weary of this transition of multiple human functions to robots. Tay’s shocking malfunction suggests that this might be cause for concern for many reasons.
- Xan Northcott