Like all human activities, AI must also set sustainability as its goal. Environmentally, the impact of giant AI development and delivery facilities is already significant, both in terms of energy consumption and water use. It is therefore necessary to reduce this impact, or AI will emerge as one of the major sectors responsible for climate change.
Social sustainability, in turn, is all to be verified. The deep fake compounds the problem of fake news with the artificial generation of images and videos indistinguishable from those shot in the real world, and this can become a problem for democratic coexistence. However, it should be considered that fake news has existed for millennia, and has always influenced society, dispensed with great ease in verbal mode: its current visual form complicates the scenario, but does not radically change it. There is another social and political risk: the ability of AI to influence people and induce them to make economic and political choices, through highly effective forms of covert persuasion. Already the spread of social networks has shown that digital media have a strong capacity for influence and persuasion enhanced by algorithms. But the biggest factor that could make AI unsustainable is its impact on employment, which without adequate policies to compensate for likely job losses, could lead to severe social imbalances.
Italiano