It is late when I enter the half-empty Sala del Consiglio and look for a place: trying not to be noticed by the speaker, I unfortunately end up in the first row, just facing myself with the smiling face of Prof. Fabio Zanzotto. I humbly sit down and try to catch up with the main topic of the conversation. I had read it was something about artificial intelligence, but now all I am hearing is about some strange kind of universal Robin Hood, a 21st century global income distributor that actually makes me think I have not entered the right room.
Every time I think about artificial intelligence, the image of human-like robots comes to my mind, in a midst of an Asimov-like fantasy and a skeptic outlook on the ever-growing power of science.
It is while I am reflecting in this way that Professor Zanzotto starts with the most disturbing part of his speech: the risk that artificial intelligence might carry is not that of a humans versus robots global war. Not yet, at least. It is the jobs market that might suffer of the harshest consequences of the increase of robots in our lives. If automation technologies will actually be able to perform big part of the tasks that humans cover nowadays, what will happen of all the people whose activity will consequently turn useless? What I called before the ever-growing power of science might turn to be the ever-present conscience of human impotence.
I have to admit that initially I have felt quite comforted by the discovery that life-long unemployment will not be a prerogative of Global Governance students. Only later the idea of a machines-led world has started to worry me. I have carried out some research, and finally discovered that, according to the McKinsey Global Institute, around 5% of current jobs are fully automatable. But, if instead that looking at jobs the focus is put on skills and tasks, the results are that more than 60% of current activities will be substantially modified or even completely revolutionized by artificial intelligence: more specifically, modified means that at least one third of the basic tasks that constitute this kind of jobs are potentially submittable to automation.
Rather than thinking of mass unemployment, it would be more precise to concentrate of mass deployment, which is the unprecedented transition from old to new modalities of work. The real dilemma is whether our societies will be able to deal with such a fast-growing rate of change in the labour market, and will be able to adapt to new jobs possibilities in sustainable and rapid way. To put in in other terms, will technology outpace society?
What is striking about the growing capacity of artificial intelligence is not only its super-fast rate of growth in terms of how many functions will be automatized, but also the method through which they can reach such an exceptional result. The real strength of AI lays not in the incredible computational capacity of modern super computers, nor in the progresses that have been made in refining the precision of robots performing tasks. It is us and our willingness to provide precious information, that makes AI the most powerful kind of technology on earth.
It seems that we should also re-orient the standard of our science-fiction nightmares. We are not in Asimov anymore: a greek tragedy would fit humanity best. Like the unfortunate Oedipus, modern societies risk to find themselves looking for the murder of the father, just to finally discover it is themselves that actually committed the crime. The provision of data, general information, even the most irrelevant pieces of it has dramatically increased the capacity of machine learning to produce actual knowledge, that is then implemented in AI processes. It is us, when searching for something on any browser, or simply by entering into our Facebook account or by watching a movie on Netflix, that provide to machine learning more and more information about our preferences, our attitudes, the kinds of service that we would pay the most for. And all this information may be more precious than the password to a billionaire’s bank account. The key element for having granted hegemony in the contemporary age is not labour, not capital, but knowledge.
The situation gets more and more tricky, as we realize to be at the same time consumers of AI, and necessary by-products to its ultimate development. The solution which is proposed by Professor Zanzotto is to carry out a strong and continuous redistributive struggle.
The very concept of redistribution has been crucial in the turning points of the last-century to enhance social rights and provide representation for growing working classes who claimed respect for their own labour. Similarly, AI users should claim for their share as contributors to AI continuous improvement.
The tragic part of this solution, is that AI users, practically speaking all of us, still don’t have any awareness of the role they play in the overall supply of services by the automation industry. It is a matter of being aware and, perhaps with a bit of courage, claim that everyone of us still does play an important role.
As the conversation is now approaching to its conclusions, I also move away from my fantasies and go back to my everyday life: classes, study, just to wait for very the moment that I go back home and watch the final episode of my favorite series on Netflix. My only hope is that is not going to take someone’s job.