In a recent article in La Nación, through the use of Artificial Intelligence, the male actors in the history of cinema were revealed, according to some factors such as versatility, technique, awards and cultural impact. The use of Chatgpt had its response by classifying these recognized interpreters: Marlon Brando, Laurence Olivier, Robert De Niro, Al Pacino, Jack Nicholson, Daniel Day Lewis, Denzel Washington and Anthony Hopkins.
“There is no definitive answer about ‘the best actor in history,’” the platform created by OpenAI initially noted and highlighted that “there are names that usually appear in any ranking serious for his career, acting quality and legacy.” Artificial Intelligence designed a list, along with a justification, mentioning the selected English-speaking male actors. In their entirety, these artists belong to the industrial heart of Hollywood, which was deployed from the late 50s to the present.
In dialogue with Noticias, the specialist in Artificial Intelligence, Elisandro Santos explained: “The response from the well-known OpenAI chatbot was diplomatic: it noted from the outset that “there is no definitive answer” to that question, but in any case it highlighted that “there are names that usually appear in any serious ranking due to their career, interpretive quality and legacy.” Instead of choosing a single name, the artificial intelligence offered a list of several figures considered among the best of all time.
“This means that the selection of “best actors” that it made is not the result of conscious reflection, but of mixing and summarizing what many critics, film buffs and specialized publications have historically said. If the majority of available sources praise certain actors, the model simply captures that trend and reproduces it. The accuracy and bias of the response depends directly on the sources: if the data it was trained on (or the pages it consults) is incomplete, outdated or biased, the response will be too,” said the expert.
Regarding this, Santos said: “Artificial intelligence models can inherit and amplify the biases present in their training data. ChatGPT has no way of distinguishing by itself whether it is leaving out actors from other regions or underestimating minorities, for example; it simply reflects the predominant information it has consumed. Recent studies have confirmed that generative AIs tend to encode prejudices and stereotypes of their sources, even going so far as to massively disseminate information that seems accurate but may actually be erroneous. In this case, the response was compound by unquestionably famous and talented names, but in another kind of consultation the bias could manifest itself in a more problematic way.
Another point of caution mentioned by the specialist, author of the book “unlock the power of Artificial Intelligence”, is that ChatGPT responses are not always reliable. The model sometimes makes errors and even “hallucinates” non-existent data,
presenting them as if they were verified facts. OpenAI warns that ChatGPT may reflect biases and stereotypes present in the data it was trained on.
For these reasons, experts recommend using these types of tools with caution: it is not advisable to take their answers as absolute truths or rely solely on them for important decisions. It is always preferable to verify information from reliable sources. For example, it would be ideal to know if the model mentioned Marlon Brando and Robert De Niro because he found many articles talking about their awards and influential performances, or if it was because of some internal statistics of how often they appear on “best actors” lists.
In conclusion, the specialist said: “Although ChatGPT was able to give us a reasonable list of great actors, we must remember that it depends entirely on its data. It has no way of evaluating the real world or updating its knowledge on its own beyond what has been provided. Its selection was an amalgam of the most cited and celebrated on the web about actors, but it lacks critical judgment beyond statistics.”
Image gallery
