Bob Dylan appears in one of three spots in IBM’s new advertising campaign starring Watson, the cognitive computer consulting unit that “enhances human expertise.” Watson can ingest huge amounts of information—legal documents, medical data, media research—interpret, evaluate and decide. Watson can recognize patterns in poetry and statistics. He can read every Bob Dylan lyric (Dylan fans know that’s a lot of lyrics) and identify primary themes as “time passes” and “love fades” all before Dylan can walk to the couch.
Pattern recognition is a survival skill, whether you are a fisherman deciding if you should eat a certain a crab or an oncologist weighing treatment decisions. But we fallible humans often detect the wrong pattern, or invent a pattern where none exists. Watson, on the other hand, isn’t limited by this inconvenient humanity. He can ingest 800 million pages in a second.
But Watson doesn’t understand sarcasm (as we see in the ad featuring Jeopardy champ Ken Jennings). And if the spot featuring Watson and a child cancer survivor tugs heartstrings, we must credit the adorable young actress—who is charmed when Watson remembers her birthday and identifies she likes things that start with “p.” He also keeps track of her medical records.
Contrary to the popular depictions of emotionally sensitive computer systems—think of Kubrick’s conniving, ultimately fearful Hal 9000 in 2001: A Space Odyssey; the loving adoptive cyber-son David in Spielberg’s A.I.; or the manipulative cyborg Ava in the British thriller Ex Machina—the IBM commercials make clear that Watson is super-intelligent, but he’s not conscious. Unlike Samantha, the beguiling operating system in Spike Jonze’s sci-fi romance, Her, Watson says he’s “never known love.” He’s also a terrible singer. You win, Bob Dylan. For now.
“I think that many of these science fiction scenarios, that computers will be like humans, are wrong.”
“Throughout history, if you wanted something intelligent, this something had to have consciousness at its basis,” said Yuval Hurari, historian and author of Sapiens: A Brief History of Humankind in a conversation with Nobel Prize winner Daniel Kahneman, professor of psychology at Princeton. “People were not familiar with anything not human, that didn’t have consciousness, that could be intelligent, that could solve problems like playing chess or driving a car or diagnosing disease.” But AI researchers have uncoupled intelligence from consciousness. Hurari goes on:
“I think that many of these science fiction scenarios, that computers will be like humans, are wrong. Computers are very, very, very far from being like humans, especially when it comes to consciousness … intelligence is a far easier thing than consciousness.”
Hurari says the great threat to humanity is not that computers will become human; it’s that computers will supersede human intelligence while carrying none of our metaphysical baggage, none of our value-less zillions of complexities. Among the most pressing complications: computers will make us economically obsolete.
In an essay for the New Statesmen, Hurari writes: “Google, Apple, IBM and other tech companies are now developing new types of non-conscious intelligence – based on machine learning and big data – that can outperform human beings in all of the above tasks and many more besides.” Intelligence, he writes, is economically valuable; consciousness is not.
Hurari offers the example of Google’s self-driving car: “A taxi driver is immensely more complex than the self-driving car,” he says. But the things a taxi driver can do that a computer cannot have no economic value. The value is in getting from point A to point B as quickly as possible.
“And when you look at it more and more, for most of the tasks that humans are needed for, what is required is just intelligence, and a very particular type of intelligence, because we are undergoing… a process of specialization which makes it easier to replace us. To build a robot that could function effectively as a hunter-gatherer is extremely complex. You need to know so many different things. But to build a self-driving car, or to build a “Watson-bot” that can diagnose disease better than my doctor, this is relatively easy.”
He concludes: “Whereas the Industrial Revolution of the 19th century created a vast new class of urban workers, the 21st century might witness the rise of a vast new class of economically useless people.”
Carl Benedikt Frey and Michael A Osborne, in their book The Future of Employment, estimated that 47 percent of US jobs are at high risk.
The great threat to humanity is not that computers will become human; it’s that computers will supersede human intelligence while carrying none of our metaphysical baggage…
With the development of AI therapy systems, the at-risk jobs might one day include psychotherapists and counselors. Fjola Helgadottir, an Oxford researcher and developer of AI therapy, writes that AI therapy systems “will administer evidence-based and effective treatments for a range of psychological disorders, but they will do so without the small talk.”
But for the time being, we don’t need to worry about clients falling in love with their therapy systems. Watson might be able to diagnose us and recommend a treatment plan, but “despite having access to a vast database of general knowledge and sophisticated natural language processing algorithms, Watson (without modification) is utterly incapable of performing any tasks outside the scope of its narrow specifications.”
Paul Thagard, a professor of philosophy at the University of Waterloo, wrote recently: “Watson for the foreseeable future lies far inferior to human abilities of dealing with perceptual representations, imagery, emotions, consciousness, learning, language, and the full range of creative problem solving that humans can accomplish. Other current AI programs share similar limitations.”
In simpler terms: “The trouble with computers is that they just don’t give a damn.”