AI Fears: Paranoia Or Astute?

Concerns about AI have been around a long time and don’t seem to be going anywhere any time soon…
For as long as there have been computers, there have been concerns that one day computers would become smarter than their creators. Artificial Intelligence, as it’s called, has been the subject of films and sci-fi novels for decades, long before each of us carried around mini computers in our pockets and on our wrists as we do now.
Indeed, over the past decade – as the pace of change in computers and digital technology has accelerated so quickly – you’d think that AI would have progressed at the same rate. But while people’s fears about AI have not subsided, it’s not clear if the prospect of computers ever possessing the multifaceted emotions and intelligence of humans has gotten any more feasible. Is AI just a luddite’s excuse for not embracing computers? Or is it an actual threat to the very human existence?
While computers and algorithms can deduce and parse a stunning amount about human behavior from our digital habits, up until now we have taught them to do that. The jump that needs to be made before computers can do this with the same complexity and nuance seems quite far off. However, big players in the tech space regularly predict that AI will one day take over. Recently, Elon Musk called AI “our biggest existential threat” while Apple co-founder Steve Wozniack was quoted telling the Australian Financial Review that “computers are going to take over from humans, no question.” Furthermore, major US institutions Bank of America & Merrill Lynch released a report that stated a “rise of the intelligent machines” would bring about “the next industrial revolution … becoming an integral part of our lives as providers of labor, mobility, safety, convenience and entertainment.”
As certain as these prognosticators sound, there are plenty of indications that AI perhaps isn’t quite the thing we need to fear. As Observer writer Charles Arthur wrote recently, to become fully human, computers would need to possess “two capacities that AI tends not to display: common sense and creativity.”
Nothing demonstrated the lack of tact that’s often found in AI better than Microsoft’s embarrassing gaffe last week with its millennial Tweet Bot, Tay. As Quartz reported, the AI program “was designed to learn from conversations with real people on social platforms Twitter, Kik, and GroupMe. But the experiment quickly spiraled and Tay ‘learned’ to be a hateful jerk in less than 24 hours online. Microsoft was forced to pull the plug and reprogram the bot.”
This was a particularly embarrassing PR stunt gone wrong for the tech giant, who was undoubtedly trying to demonstrate this advancement in AI technology but which immediately backfired. The bot quickly spewed racist, offensive and misogynist Tweets because it didn’t possess the common sense to know that just because some people speak that way, it doesn’t mean it should. It could only speak what the data it ingested taught it, it couldn’t filter what it learned through the norms of human behavior. The episode was a perfect example of why computers, left to their own devices, simply don’t possess many of the qualities that are integral to being a sensitive human.
With time and development, it is possible that AI will be part of our lives, but the prospect of it taking over the world without humans being able to thwart it seems sensationalist. Technologists are currently investing research and money into how to use AI for beneficial reasons, including initiatives like OpenAI.org, which is an NGO that’s creating an open-source form of AI that is accessible and available to everyone, not walled within a major tech giant.
It remains to be seen what will happen to AI, but one thing’s for sure: if it can’t figure out how to Tweet, it’s likely not going to surpass the sum total of human intelligence any time soon.

Explore VPS’s intelligent hosting solutions through our managed services.