Carlo Ginzburg is a well-known historian, according to some statistical measures one of the most famous living historians:
https://pantheon.world/profile/occupation/historian
He was for many years a Professor at UCLA and he is now living in Italy, where he has been affiliated with the University of Bologna and the Scuola Normale of Pisa. He also has direct personal connections with two topics we will be discussing in this interview.
Dear Professor Ginzburg,
One of your grandfathers was Giuseppe Levi, a biomedical scientist mentioned by 3 Nobel Prize winners, Renato Dulbecco, Rita Levi Montalcini and Salvador Luria, as a mentor that played a key role in their scientific careers. We present here some of their memories about Giuseppe Levi and about a time when him and a few others acted as moral leaders for many younger scientists.
Carlo Ginzburg:
One of the most striking stories about my grandfather, described in the page of historical sources you mention, is that told by Rita Levi Montalcini, whom I have also met. I am referring to his rapid shift from a strong criticism of some of her ideas to encouragement and approval after being presented with data and evidence. This behavior is not so common among senior authorities in science (and indeed also in humanistic disciplines). It was attributed by Rita and others to my grandfather's passion for science, which he put before a desire for personal honors and praises. It was indeed one of the ways for younger scientists to see that his commitment was sincere, and it represented for them an inspiration.
Another of your personal stories we found interesting, and possibly relevant to our discussion about AI and science, was your description of a personal change in some aspects of your identity. You have told before how you vividly remember that, as a child in Italy during the second World War, you were with your mother and grandmother at a place occupied by the German army. Your maternal grandmother, who was not Jewish, told you not to use not your father's but her own last name (obviously, in order to avoid being identified as a Jew). You said that retrospectively you realized that at that very moment you became a Jew. This represented a wider phenomenon in Italy at the time: for example, Primo Levi has said that "They made me become Jewish. Before Hitler I was a middle-class Italian boy" (1). A well-known postwar Italian writer (Elsa Morante) made a similar statement. You have also stressed that all of us have multiple group identities and that you do not like the use of this concept (“identity”) as a weapon to separate insiders from outsiders and marginalize either groups or individuals. The analogy this story suggested to us is that the challenge posed by AI for human understanding could strengthen our identity as humans. But it might be a simplistic analogy.
Carlo Ginzburg:
All analogies are simplifications and arguments might be presented for both sides even in this case. I will play the devil's advocate, in my attempt to summarize our discussion.
Some of my reflections about human identity have been inspired by an essay by Italo Calvino, where he wrote: "Identity is therefore a bundle of diverging lines that find in the individual their intersection" (2). You refer to my story as a moment of change in my identity, but the change was of a relative nature and the complexity described by Calvino is still a good description, even for myself.
We also discussed the book by Benedict Anderson on imagined communities (3), which analyzes a widespread type of group identity, i.e. national identity. National identities are the outcome of a social, cultural and political construction, not natural entities: this is why Anderson calls them “imagined”. You suggest that a similar process, based on AI, could reinforce human identity, but other group identities are not likely to disappear.
We know from Darwin that biological evolution depends on variation and Kuhn has shown how new scientific paradigms can replace old ones, starting from ideas initially held by a few innovators (4). The exchange of ideas among biomedical scientists you describe will presumably pave the way to fruitful developments, but I would refrain from connecting AI to another identity. I would regard it as just another technology: a tool that can be good or bad depending on the use we make of it, like the internet. If AI will be able to produce original conceptual advances, the challenge to human collective intelligence will be real and will produce a constructive human reaction, but would it reduce the amount of social inequalities across the world? This seems unlikely: the analogy provided by the internet is telling enough. After all, the very idea of “human collective intelligence” implies (and conceals) inequalities.
In my field, history, looking at queries to ChatGPT made in Italian, I am aware of quite disappointing results, but things might differ in scientific fields. In any case AI is clearly improving rapidly now and we must be ready to follow its developments and evaluate them with an open mind.
REFERENCES
1- Levi, Primo. "Conversazioni e interviste 1963-1987" (Interview with Edith Bruck) Einaudi (1997).
2- Calvino, Italo. “Identità”, "Saggi 1945-1985" vol. II, pp. 2823-27 Mondadori (1995).
3- Anderson, Benedict. "Imagined communities: Reflections on the origin and spread of nationalism" Verso books, (2006).
4- Kuhn, Thomas S. "The structure of scientific revolutions" University of Chicago Press (1962)