Enormous investments have been going into artificial intelligence for some time now. In some cases the returns have been unclear; neither are we entirely sure what artificial intelligence, machine learning and algorithms are, or what the difference between them is. What is certain, however, is that the future of the economy and of the society is inextricably bound to this new technological revolution. Capital amounting to 15 billion dollars has been poured into 2250 start-ups in the United States, many of which will fail, and some estimates put global spending on artificial intelligence at 13 trillion dollars by 2030.
It’s enough to consider some examples in daily life of the transition from automation to autonomy, such as the driver-less car. It has been calculated that 21 sectors of the economy will be affected by the adoption of the autonomous vehicle, with not only its technological implications to consider but also those legal, regarding who will be managing the data, and ethical regarding fault in the case of accidents.
Data management helps complex corporations form a comprehensive vision of customers and their needs. An age-old concern, in reality, going all the way back to, data management remains central not least because of the massive quantities involved: 1820 terabytes of data created per second as a result of telephone and mobile use alone; and a full 200 billion e-mails sent every year. Since the energy used to send one e-mail is equivalent to a 60-Watt lightbulb burning for 25 minutes, the environmental issue is not only one of saving paper but also of energy consumption.
Approaches to regulation are highly diverse. Europe has chosen a route that China and the United States have steered clear of; for the EU this could mean a way to forge a common framework for defense. At the same time, the geopolitical and economic interface between China and the U.S. is leading to a technological cold war and probable permanent structural tension between liberal democracy and digital authoritarianism.
There was much discussion of the value and power of algorithms, which are surely a great help in strategic sectors such as health, but are also rife with ethical, economic and political problems. Algorithms are increasingly key to decisions in sectors such as banking, for example, when it comes to approving mortgages and loans. For optimists, this avoids a human manager’s potential errors in judgement, but for pessimists can lead to fatal errors precisely because the applicant’s “human” reliability is not taken into consideration – a certain something that surely eludes the power of data, even though in some cases results can be based on 30 million parameters.
Humans having the final say was a notion that underpinned the entire debate on artificial intelligence. The human brain – which processes a billion billion operations per second – is not about to be replaced: algorithms are not enough, we still need humans. Thus technology must be applied to the idea of economic sustainability. With the knowledge currently available to us, it would take a very large room to house something that resembles a human brain. So for the time being robots, which are very useful in both industry and in private, will remain “stupid” and interlinked by cloud: many bodies, one brain.
Furthermore, technologies are marketable only if they are cost effective, which means if they are applied to complex structures to resolve complex problems. For instance, the productive use of artificial intelligence, machine learning and algorithms in the healthcare sector could mean two GDP points for a country like Italy. Technologies change rapidly however and can result in just a few short years in obsolete jobs and instruments, which is why the public and private sectors must agree to keep abreast of these sudden mutations. At the same time, ethical mindfulness is going to have to accompany developments in digital technology and artificial intelligence in order to be able to recognize what is socially and politically applicable.