How to navigate the changes ushered in by the use of Artificial Intelligence (AI)? The year 2023 will be remembered as the year that AI “took flight” – an event that has raised some urgent questions regarding the need to develop and regulate that technology and train people to use it as a resource and not as a menace. Statistics on the adoption of AI systems such as ChatGPT show how useful the society considers these instruments, but can this success perhaps mean that too much human thought is being entrusted to AI?
The question is a metaphysical one since, although AI is “intelligent”, it is also devoid of “responsibility”, which is one of the cardinal prerequisites for the definition of a “person”. Moreover, AI’s ethical paradigm must be controlled accurately, as some practical experiences have shown, where proposed solutions to urgent problems – such as climate change – are unacceptable from an ethical standpoint: an example is the AI proposal to bomb the planet’s 15 major cities so that the resulting dust would filter the sun’s harmful rays.
At the same time, the utility of AI in business and computational processes has been widely confirmed, yielding a substantial reduction in design timeframes and resources required; a fact clearly acknowledged by the 61% of Italian firms that have already launched AI initiatives. That percentage, however, drops to 15% among SMEs. In any case, this is a market that has seen exponential growth and was valued in 2022 at 500 million euro.
One of AI’s assets is its capacity to automate compilation and sorting processes, thus changing the way the accounting and legal services sectors work. Bookkeepers, in fact, are now able to concentrate more on consulting and other value-added services as opposed to merely preparing and filing tax forms, which can now be done by a machine.
The limits of certainty in the use of AI can be pinpointed in the legal sector. Automatic Judgement experimentation is underway in programs such as Lithuania’s “Justitia”, which passes “automatic”, yet appealable, judgement in cases of minor value. Nevertheless, this raises doubts not only on the possibility but also on the prudence of a “robot judge” capable of deciding human issues.
Thus, caution must be used in introducing Artificial Intelligence, given the plethora of studies on the application of various technologies based on purely mathematical-statistical methods with no unified definition. AI is to be parsed in function of the specific context, since the discipline itself assumes connotations entirely different when implemented in the recognition of texts or the generation of new content.
The sector of art is also being affected by AI, with some creations resulting from the replacement of traditional composition processes with prompt-engineering. Yet, it is widely felt that the creative process – at least in the case of high-level works – is complex precisely because the human factor represents the basic underlying feature of the concept of creativity, and can hardly be substituted with the sophisticated plagiarisms of AI.
It would therefore by advisable to ensure that the spread of the society’s use of AI be accompanied by the development of critical thinking, not least because the risk is that the continuous simulation, emulation and revision of reality could render reality itself superfluous.
In this sense, three essential measures are needed: the completion of training processes so that the aware use of AI becomes part of every study plan, not least with the aim of avoiding the digital divide that the technology has already created; adoption of a legal regulatory structure that ensures that AI is adopted for the benefit, and not the exploitation, of humanity; the introduction of fact-checking systems capable of verifying whether photos, videos and news reports are “real” and not AI-generated.
In order to avoid wasting the major development opportunity offered by this technology, it remains necessary to pursue forms of regulation and development. An ideal project could be to gather forces at European level with a view to developing a sort of “AI CERN” built on a public/private partnership, and that would be capable of offering a successful and avantgarde model at global level.