I recently came across a quote by the father of Information Theory and MIT Professor Claude Shannon. “I visualize a time”, he said, “when we will be to robots what dogs are to humans, and I’m rooting for the machines.” Shannon did not seem to like human beings too much, but this quote set off another thought process in my mind. As a technology writer and digital transformation practitioner, the second most asked question to me is: “Will AI take our jobs, and what should I do to protect mine or my child’s”? (For the most oft asked question, you will have to read on to the end of this article!).
Whether AI will take our jobs or create new ones is one of the greatest debates of the modern world. It is not a question that has been asked for the first time. Every instance when a revolutionary new technology comes in, the same thought fearfully raises its head. It bothered Ned Ludd in 1779 after the Spinning Jenny was invented, which threatened to take his job as a textile factory apprentice. He went and smashed a machine or two, catalysing a movement against textile technology, and started the Luddite movement. New age Luddites worried about PCs and their job-destroying potential, this movement was particularly strident in India with computers being smashed by workers unions. It turns out that the IT revolution created millions of new jobs, and catapulted India into a tech superpower.
But AI, everyone says, is different than all these technologies, and we will need to think about it in a different way. They may be right, this is the first time that there is a technology which potentially replaces us, and perhaps could be even more powerful than us, and theoretically has the potential to make humans redundant. Turing Award winner, Alan Perlis articulates this feeling best when he says, “A year spent in artificial intelligence is enough to make one believe in God.” Most global institutions, though, are much more sanguine about AI. The World Economic Forum predicts that by 2022, AI will create 133mn new jobs, though COVID might impact that prediction a little bit. It also says that from a 30/70 ratio of division of labour between a machine and human will dramatically shift to become 52/48 in 2025. IT consulting firm Garner claims that AI-related job creation will reach two million net-new jobs in 2025.
But where will these net-new jobs come from? I have a simple way to think about it: most of the work we do can be rudimentarily divided into English (or any other language), and Arithmetic. English is the creative part – strategy, communication, messaging, while arithmetic is the analytical part –excel sheets, number crunching, financial planning. In my simplistic thinking, the arithmetic bit will get taken away by robots first (case in point: RPA, or Robotic Process Automation) and the humans will still own the English bit. However, when the English part has already start getting chipped away by AI, specifically Deep Learning, Neural Networks and now GPT-3. AI now creates great music, search for AI written music in YouTube and you will find thousands of them. Microsoft Research and ING teamed up to have an AI paint a Rembrandt painting (nextrembrandt.com), and it painted a critic-defying one, six hundred years after the master was dead. AI writes poetry and prose, and defeats humans in games which are instinct and imagination driven – the epochal defeat of World Go champion Lee Sedol by AlphaGo.
Perhaps the best explanation of the AI and jobs is given by Kai Fu Lee, the acclaimed AI investor and practitioner. His famous matrix has optimization-to-strategy on one axis, and no compassion-to-full compassion on another. High optimization – low compassion jobs (like telesales, customer support, dishwashing, radiologist, truck driver, etc.) will be the first to go and high compassion- jobs like CEO, M&A expert, teacher, elderly care worker, etc. will be last. There will be AI-only jobs, jobs where AI will support humans, and jobs where humans will assist AI. Then there will be jobs which will always be there for humans, and those will have the need for communications skills, empathy, compassion, trust, creativity and reasoning, in fact, he has a ready-reckoner list of the ten safest jobs: Psychiatry, Therapy, Medical care, AI-related research and engineering( ironically, more AI requires more AI engineers and researchers), Fiction writing, Teaching, Criminal law, Computer science and engineering, Science, and finally, Management.
If you think it, it is more about humans than it is about AI. Udacity co-founder Sebastian Thrun puts it best, “Nobody phrases it this way, but I think that artificial intelligence is almost a humanities discipline. It’s really an attempt to understand human intelligence and human cognition.” And that, naturally brings me to the question which people ask me most, “Will AI replace humans?”. The short answer is that in some ways it will, for the long answer, we will need to wait for another column.
(This article was published as an OpEd in Mint dated Oct 16, 2020)