The promise of artificial intelligence seems to be all the rage today. The media is full of wide-eyed discussions about the promise of AI and how this is going to transform economies and societies. The technology has certainly moved from the labs into every day lives. It is currently being used in medical devices, smart-home systems and video games. IBM is aggressively marketing Watson as “working alongside people in 45 countries and 20 industries to help businesses work faster and work smarter.”
The most compelling example is the use of AI in driverless cars which many predict will be the way of the future in about 15-20 years. Medical students are now wary of choosing specializations like radiology and pathology because AI is extremely good at reading images and slides. For the past several years, the most lucrative application of AI has been in online advertising. AI is expected to take over routine managerial work like coordination and control tasks, scheduling and reporting. In a recent interview with Wall Street Journal, Andrew Ng, the co-founder of Coursera and former AI scientist at Google and Baidu says that “AI is the new electricity. Whatever industry you work in, AI will probably transform it, just as 100 years ago the rise of electricity transformed industry after industry – everything from transportation and communications to manufacturing to health care.”
What passes of as Artificial Intelligence is neither “artificial” nor really “intelligent”. While Deepmind can beat humans on the Go board and Watson beat the best competitors at Jeopardy, ask each to play Family Feud and they will get stumped. It is human ingenuity (and intelligence) that enable these AI machines to be trained and become very good at specific tasks. While there has been much progress in enabling AI machines to train themselves in many problem domains, humans have to create the algorithms and the rules that actually gets these “robots” to solve problems.
The fact is that AI machines are unintelligent and really cannot deeply think for themselves no matter what their proponents say. They certainly cannot reflect on society’s values and priorities. Consider this: Recently, there has been increasing talk on the U.S. using AI machines to fight wars. What could be better? One can inflict significant damage on the enemy while minimizing loss of lives on our side to the point of not having any soldier in the theatre of war. Every country in the world believes that unarmed civilians should not be killed in a war, and certainly women and children should not be harmed at all. No autonomous AI war machine can come up with these values on their own. What’s more, these intelligent robots would also not be able to learn values even after many trials and errors. The drones that the U.S. used under the Obama administration to destroy the Afghan militants on the Pakistan border ended up causing such civilian casualty that their use had to be suspended. No matter how clever the programmers were in training the drones, these non-human machines could neither imbibe the values well established in the U.S. and elsewhere nor even act on them properly. The Campaign to Stop Killer Robots urges governments to heed an open letter signed by 126 founders and directors of more than 100 robotics and artificial intelligence companies from 28 countries demanding urgent action to address concerns of having fully autonomous weapons.
People who create the new machines based on artificial intelligence are very clever problem solvers. Need intelligent machines to read x-rays and pathology slides correctly, done! Need robots to fight fires in large complex building sites with minimal collateral damage, done! Need cars to drive themselves long distance and avoid accidents on the highway, done! Armed with STEM degrees these problem solvers from the California Silicon Valley to Manhattan’s Silicon Alley and Austin’s Silicon Plateau are working very hard to use intelligent machines to solve complex corporate and societal problems effectively. And digital problem solvers are being rewarded handsomely by the global capitalist system that dominates the world today.
However, for this brave new world to be one we really want we need value creators as much as problem solvers. It is very important for every country to set parameters and boundaries on what is socially acceptable and what reflects the value system prevalent in society. Value creators must be in the middle of this conversation and debate. They will bring empathy, ethics and engagement with society to the technologists who are trained in solving problems. Value creators will also help define what problems need to be solved. Do we really need autonomous vehicles on U.S. highways? Do we need to replace the human touch and empathy of doctors with Watson conveying the diagnosis of an illness to a patient? Should we invest money in intelligent technology that extends the lives of 90 year olds rather than saving the life of an infant? There are no ready answers to these questions. However, trusting those excited and excitable by pure technological advancement and very good at solving technological problems to provide answers to these value laden questions would be quite foolhardy.
We need to nurture and reward value creators educated in the social science and humanities to work hand-in-hand with STEM centered problem solvers to create a better world.
Let me end with two quotes that should really awaken those mindlessly pursuing technological advancement. Sun Microsystems cofounder Vinod Khosla apparently is of the view that “Little of the material taught in Liberal Arts programs today is relevant to the future”, a view held by a vast majority of techies in Silicon Valley. This should really give us pause. Let’s not forget Robert Oppenheimer, the leader of scientists and engineers who worked hard to enable the atom bomb, quoting the Bhagavad Gita “I am become death, the destroyer of worlds.”