Web Science and Digital Libraries Research Group

Figure 1: A sample prompt used in generating persona using ChatGPT (Figure 1 in   Park et al. ) Introduction: The Unseen Biases of Our AI Companions Large Language Models (LLMs) such as ChatGPT have become ubiquitous in our lives. We rely on them as impartial conveyors of facts for everything ranging from composing emails to solving complex queries. However, if we put these AIs in the hot seat and question them about identity, which is a very human thing, what would be the outcome? What if we ask them about autism? In the recent study “ ‘As an Autistic Person Myself:’ The Bias Paradox Around Autism in LLMs ” ( CHI 25 ), Park et al . went deep to understand what ChatGPT thinks about the condition of being autistic. The results they got weren’t just a malfunction of the system; rather, they were an indication of the way people behave towards the issue of neurodiversity,  going back and forth between one extreme and the other. The artificial mind is going through the so-called " ...

Figure 1: Mementos of Jack's first tweet, archived at different points along the Twitter UI timeline. In the animation in Figure 1, we can see images of four archived versions, or mementos , of the first public tweet, posted on March 21, 2006 by Jack Dorsey (co-founder of Twitter). However, each memento has a different user interface (UI). This is because the UI of Twitter has changed over the years and these mementos were archived at different points in time. The widely popular microblogging platform Twitter (now rebranded as X) was first launched in March 2006 , and since then there have been many changes in its UI, including how individual tweets are rendered on the desktop web, which we refer to as the tweet UI . The mementos in Figure 1 show noticeable changes to the tweet UI. Changes to the tweet UI, though subtle to the human eye, can have a significant impact on the performance of certain research tasks. A model trained largely on a dataset of images displayed in the lates...