AI Proposal Paper Topic:

My proposal idea/topic for Assignment 1 will be to research how to ChatGPT (a multimodal generative AI system using large language models) “works”, it’s interface,  as well as analyzing potential biases and how clerical human error can be harmful when programming these systems of artificial intelligence (real world/social scientific impact). I will also include how the diversification of who’s in tech (ex: currently dominated by white men) can create a positive impact on how inclusive highly advanced chat modules have the potential to become (ex: a team of black woman may produce more accurate results about black hair care, vernacular, and culturally specific values than a team composed of those who have not been exposed to the same).

Below is the same proposal idea reworded by my chosen AI tool, ChatGPT:

My proposal idea/topic for Assignment 1 will be to research how ChatGPT, a multimodal generative AI system that uses large language models, “works,” including its interface, training, and ability to generate responses. I will also analyze potential biases within AI systems and how human decision making, errors, and assumptions during the development and programming of these systems can have harmful real world and social scientific impacts.

I will further examine how diversification within the technology industry can potentially create more inclusive and representative AI systems. For example, having people from different racial, cultural, and experiential backgrounds involved in the development and evaluation of AI may help identify biases and gaps that others may not recognize. This can be particularly relevant when examining culturally specific subjects such as Black hair care, vernacular, and other forms of cultural knowledge that may be underrepresented in technological systems.

Using sources such as Kate Crawford’s “Artificial Intelligence’s White Guy Problem” and Ruha Benjamin’s Race After Technology: Abolitionist Tools for the New Jim Code, I will examine how representation, data, and human decision making can influence the development of AI and potentially reproduce existing social biases.