Artificial Intelligence - The Good, Bad and Ugly
Behind the daily tech headlines, familiar names such as Amazon, Google, IBM and Microsoft are quietly engaged in a platform war. Artificial Intelligence (AI) is about to enter the mainstream, and when combined with machine learning capabilities that are constantly devouring human knowledge, it appears that the battle for information control is already underway.
The Good
We are repeatedly subjected to man-versus-machine comparisons to illustrate progress. Rather than being dazzles by IBM Watson destroying human competitors at Jeopardy or Google’s AI beating the Go world champion, the real race is to deliver the first Amazon of data. These familiar names have been busy adding machine learning capabilities to their cloud software platforms to ensure that they continue to dominate the tech industry.
Twenty years have passed since the IBM super computer Deep Blue beat the world chess champion, Gary Kasparov. The quest towards artificial intelligence continues as Facebook, Microsoft and Google are all investing heavily in deep learning technologies
Over a third of people believe AI poses a threat to humanity, which has probably come from watching fictional movies such as the Terminator franchise. In their defense, with each passing year the line between fantasy and fiction becomes ever more difficult to differentiate. Meanwhile, exaggerated claims by the media continue to plant the seed of fear.
Microsoft’s proof-of-concept communication tool called holoportation allows users to teleport a 3D hologram of themselves to anywhere in the world. Were this technology to take off, it would undercut the major criticism of communication technology and possibly social media, which is that it detaches people. A holographic chat is surely better than a text message. The way I see it, this is not technology to fear, and pays tribute to the good intentions of many creators around the world.
The Bad
That doesn’t mean we shouldn’t exercise caution. These are uncharted waters. The applications of much of this technology may extend beyond what its creators initially intended. The unintended consequences warrant consideration.
Bill Gates and Elon Musk have warned about the potential perils of AI being unleashed on the world. At Vanity Fair’s New Establishment Summit, Elon Musk advised, “Most people don’t understand how quickly machine intelligence is advancing. It’s much faster than almost anyone realizes, even within Silicon Valley; and certainly outside Silicon Valley, people really have no idea.”
In the midst of this opacity it becomes problematic to come at pros and cons. It is also unclear what, exactly, the possibilities of AI are. This highlights exactly why there is so much confusion around the subject. Much of what we hear and read is conjectural, and nothing more. In any event, we can be sure that despite being programmed with a wealth of useless information, it seems AI struggles with any form of empathy.
Take as an example Microsoft’s experiment with real-time machine learning, Tay. Tay was allowed to learn from Twitter users in real-time. It ended in humiliation when the bot quickly learned the darker side of Twitter and began using sexist and racist language to produce incredibly offensive tweets. Microsoft:
We are deeply sorry for the unintended offensive and hurtful tweets from Tay, which do not represent who we are or what we stand for, nor how we designed Tay.
Recent events during the infancy of AI and machine learning are a cautionary tale. They also underscore the complex differences and similarities between man and machine. There is an argument that we should program these machines for enforcement given that Tay learned the nefarious side of Twitter so efficiently.
The Ugly
The ugly is less about aesthetics and more about order. This is a messy affair, especially if programmers and creators attempt to replicate the more abstract aspects of human thinking and behavior.
Certainly a machine beating a human at a game is impressive, but humor, sarcasm, satire, micro-expressions, eye contact, or general body language are quite different. They have obscure parameters and are awash in nuance. For these very reasons it is proving to be quite troublesome.
Despite the advances in AI, technology-savvy investors remain skeptical about overhyped gimmicks. The implications for society are unclear, and as more and more specialized work is turned over to automation, who will be left without a job?
The optimists among us say that history should convince us that humans would weather this storm and not lose themselves in the process. Others still would say that were we to lose high-level specialized work, it would free up our time for creative and leisure activities. The question remains: How will we support ourselves financially when jobs grow slimmer and slimmer?
Conclusion
We are in the middle of a digital transition that cannot be stopped. I remain hopeful that these advances in machine learning and AI will deliver positive change to society, that we won’t lose our sensitivity for collateral damage, and that we won’t succumb to the temptations of technology for technology’s sake.