Artificial intelligence represents one of the most fascinating achievements in contemporary innovation
We listened to the first three episodes of GenIAle, Matteo Bordone's podcast for il Post and FAIR, and three concepts in particular stuck with us.
The first aspect concerns the concept of intelligence. As we know, the long-term goal of AI is to simulate human intelligence, achieving the status of Artificial General Intelligence (AGI), as it is called in technical jargon. A much more complex ambition in practice than in theory, given the level of advancement and differentiation of human intelligence. Just think of the famous Emotional Intelligence by Daniel Goleman (Emotional Intelligence, D. Goleman, 2011), an empathic capacity that to date seems to belong only to human beings.
The complexity of our brains does not stop there, however, since like the emotional one, there seem to be still other types of intelligence, as many as eight, as psychologist and Harvard professor Howard Gardner claimed in his book Frames of Mind published in 1983. In short, as much as the results of the latest AI tools are starting to become very advanced, there is still a long way to go to reach the levels of human intelligence. Much training is needed to get to the creativity, deduction and other characteristics of our intellect.
This awareness of the potential of human intelligence should help us see with less fear or dread the equal potential of artificial intelligence, and here we come to the second point. Since the release of Chat GPT3.5 in November 2022, there have been all kinds of articles about how technology will replace humans in an apocalyptic scenario worthy of the best Will Smith in "I, Robot."
While, as recently reported in a post by Will media, some jobs will indeed be less in demand (due to a number of factors including AI), on the other hand the tools at our disposal can be of great help to people and companies in everyday tasks, or even in reinventing themselves in new jobs. Even greater still seems to be the future potential in fields as fundamental to us as research in medicine and energy. During the Italian Tech Week 2024 in Turin, John Elkann (CEO of the Exor Group) and Sam Altman (CEO of OpenAI) were discussing, for example, how quickly, by 2050, coal might finally be replaced, or the new frontiers of interaction we might have in our daily lives with transportation and many other services we use.
As frightening as new things can always be, accepting innovation by trying to understand how certain technologies work and how they can be useful to us is the first step to take in order not to give in to agitation. And here we come to the third point: approach AI in the right way. The AI tools at our disposal are evolving very rapidly, achieving substantial improvements between versions. Just think of the difference between Chat GPT3.5 and the new Chat GPT 4-o, in terms of output and reliability of responses. This does not detract from the fact that we must always be careful about the answer we are given, since we are not dealing with an absolute oracle, but with a very advanced system that stores information and returns it in the form of words that have a high probability of being found one after the other.
This confronts us with two main risks, the bias and the hallucinations. I bias occur when the tool collects information that is biased in some way in the sources, and therefore will give an equally biased response. For example, it is very common for tools such as Google Translate, which have always used AI, to have gender biases in many translations, especially in the field of business. The hallucinations on the other hand, occur when responses stray too far from the underlying theme and meaning of the prompt entered, precisely because it is not a thinking mind but a probability game between consecutive words within a sentence.
Our conclusion is this: artificial intelligence represents one of the most fascinating achievements of contemporary innovation, the result of human creativity and genius, capable of opening up new frontiers in various fields. However, in order to fully exploit its potential and avoid any risks, it is essential to thoroughly understand its mechanisms and limitations. Approaching AI with awareness and openness allows us to reap significant benefits without giving in to fear, increasingly valuing our unique and complex human intelligence.
_