For laudators, not a single product, not a single advertising brochure, not a single service is mentioned without mention of “machine learning” or algorithms. These same speeches are generally embellished with a few magic words or “buzzwords” such as “deep learning”, “reinforcement”, “frugality”, without forgetting of course the notions of ethics and trust, which have been particularly fashionable lately.
Artificial intelligence (AI) even comes to be “personified” even if it is, although extremely complex, only a technology: AI becoming the alpha and omega of our existence, the dream answer to all of our problems. The major advance represented by the development of AI indeed suggests almost infinite possibilities. Productivity gains are already being felt for companies that have adopted this technology. McKinsey estimates that by 2030, AI could bring an additional 1.2% annual growth worldwide.
The use of Artificial Intelligence tools and techniques opens up new perspectives in many fields such as education, industry, transport, agriculture and many others.
“Making sense of Artificial Intelligence”
The detractors meanwhile navigate between “the supposed dictatorship of algorithms”, the great replacement of man by machine and other dystopian theories. More down to earth and far from theses and dystopian universes, the use of algorithms for purposes of surveillance, reorganization of work or even the energy impact of certain technologies, in reality raises much more legitimate questions. Fears about AI are also at the center of the 2018 Villani report, “Making Sense of Artificial Intelligence”. Cédric O, regularly raised the issue of the digital divide, reflecting a real concern even in the highest spheres of the State.
Although offering tremendous opportunities, AI is accelerating the rise of digital in our societies while making it more complex to master. In such an ecosystem forged by algorithms, only users with a certain level of knowledge and mastery can keep their free will, understand their environment and not let themselves be locked in a bubble.
Let us nevertheless agree that between laudators and despisers there is a path which, while recognizing the contributions of technology, does not conceal either the biases or the threats that certain uses can produce. Let’s not throw away all the contributions of AI in medical imaging and the early detection of cancers on the pretext that generalized facial recognition can be a danger to our freedoms, any more than we should question part of our freedoms to facilitate the fight against money laundering.
“Science without conscience is nothing but ruin of the soul”, it was already said some five hundred years ago.
Even if many organizations, scientists or members of civil society are mobilizing to make the ethics of AI a central point, should we, for all that, leave this responsibility to the sole discretion of insiders or don’t we have to democratize knowledge, to ensure that the understanding of technology, its capacities and its limits becomes the fact of all and that AI is neither divine nor the incarnation of the demon.
AI is, rightly, an essential lever of power… provided that this technology is accessible to all spheres of society and is not the prerogative of only a part of the population. Thus, preventing the development of digital divides is a decisive issue. Artificial intelligence is therefore a phenomenon that should be addressed in its various aspects: to focus only on the opportunities it brings is to hide the setbacks it can cause. The ability of States to adopt a strategy to develop their technological level while ensuring satisfactory digital education for their population will be decisive in order to benefit from AI without suffering its setbacks.
At a time when science is sometimes decried, when conspiracy theories flourish on purpose or through ignorance, let us ensure that in all our schools from primary to high school, from universities to major scientific schools, our young generations are educated just as much about benefits than risks. Let’s make sure to explain, educate and demystify for everyone what this technology can bring and make everyone understand that, as Kate Crawford so aptly puts it:
“AI is neither intelligent nor artificial”.