As mentioned in a previous post, at the beginning of the month the Collins English Dictionary announced that AI (Artificial Intelligence) was the word of the year. Their shortlist was pretty exciting, though, as it featured a wide range of terms, from "greedflation" ("the use of inflation as an excuse to raise prices to artificially high levels in order to increase corporate profits"), "nepo baby" ("used to describe the sons and daughters of celebrities whose careers are assumed to have taken off thanks to their famous parent") and "deinfluencing" ("the use of social media to warn followers to avoid certain commercial products, lifestyle choices, etc").
The shortlist of terms also included "semaglutide", the appetite-suppressing medication, the acronym "Ulez" (for "ultra-low emissions zone" referring to an area in which only vehicles that emit very little pollution are allowed to enter without paying) and "canon event", a definition taken from the animated film "Spider-Man: Across the Spider-Verse" and indicating an event that is essential to the formation of an individual character or identity.
AI, described by the Collins English Dictionary as "the modelling of human mental functions by computer programs" has been much talked about in 2023: on this site, for example, we posted over 40 features about Artificial Intelligence in connection with art, fashion, medicine, science and law (between December 2022 and November 2023).
The annual word of the year is usually selected by lexicographers monitoring a range of sources, including social media, according to the publisher. Word year of 2022 was "permacrisis", while "NFT" was chosen for 2021 and 2020's word of the year was - obviously - "lockdown".
But if the use of the word (strictly an initialism) quadrupled over the past year, not all agree with this choice. For the Cambridge Dictionary the word of the year is indeed "Hallucinate".
Though the verb may evoke in our minds acid trips and drug-induced states, Cambridge Dictionary chose it as word of the year in reference to AI. According to their definition: "When an artificial intelligence (= a computer system that has some of the qualities that the human brain has, such as the ability to produce language in a way that seems human) hallucinates, it produces false information".
In many ways it is a great word as it perfectly summarises the fears and anxieties surrounding AI, the process of asking something to the AI and obtaining false information. It is indeed important to be aware of the potentialities behind AI, but also of its weaknesses.
"As excitement meets reality, the limits of generative AI tools are as fascinating as their potential," Cambridge Dictionary explained about its choice. "AI hallucinations remind us that humans still need to bring their critical thinking skills to the use of these tools. Large language models are only as reliable as the information their algorithms learn from. Human expertise is arguably more important than ever, to create the authoritative and up-to-date information that LLMs can be trained on."
On the Cambridge Dictionary you will also find other AI-related entries this year, including large language model (or LLM), generative AI (or GenAI), and GPT (Generative Pre-trained Transformer), yet "hallucinate" is the most interesting one as, while it pertains to the semantic field of psychology, it is nowadays used more in the technology field.
In a way we could also start applying it to human beings behaving like an hallucinating AI: it wouldn't indeed be improper to say that those politicians lying about the state of their country while things are far from perfect, do hallucinate like an Artificial Intelligence system producing false information.
While AI triumphed in language polls this year, there is currently a drama unfolding in the world of Artificial Intelligence and technology: news broke that yesterday OpenAI fired its co-founder and CEO, Sam Altman, reportedly for lying to the company. According to a statement from the board, "Mr. Altman's departure follows a deliberative review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities".
Altman played a key role in founding OpenAI in 2015 as a non-profit with a $1 billion endowment from high-profile supporters (including Elon Musk, CEO of Tesla and X, the social network formerly known as Twitter, Peter Thiel, co-founder of PayPal, and Reid Hoffman, co-founder of LinkedIn). The company shifted to a "capped profit" model in 2019, with Altman as CEO.
Altman has been at the forefront of efforts to develop "artificial general intelligence" (AGI), capable of completing human-like tasks and, last year, after OpenAI successfully launched ChatGPT, Altman, who started getting known as the "father of ChatGPT" and the "Oppenheimer" of our age, gained prominence as a visible tech executive and an AI expert.
Throughout this year, though, Altman displayed an ambivalent attitude, urging the members of US Congress to regulate AI, but then criticizing the EU on forthcoming AI regulations which he thought excessive. In September he testified in Washington at the AI Insight Forum, and in November he went to the UK's AI safety summit. Mira Murati, OpenAI's CTO, will take on the role of interim CEO.
Following his departure, Greg Brockman, the President & Co-Founder of OpenAI, also resigned from his role as board chair (OpenAI originally stated that he would remain in his role with the company). Brockman and Altman also issued an official joint message on social media to explain what had happened.
The news triggered a series of resignations within the company, including Jakub Pachocki (OpenAI's director of research), Aleksander Madry (head of a team assessing AI risks) and researcher Szymon Sidor. The specifics of what Altman allegedly concealed from the board remain unclear, prompting online speculation about potential additional reasons behind Altman's dismissal, possibly related to AI risks.
It appears that the year has definitely been marked by significant developments of all sorts in the field of Artificial Intelligence - and it's not over yet...
Comments