Cafe Inquiry

This month's discussion topic: The Impact of Artificial Intelligence

Recently AI has been in the news with the promulgation of large language learning models and art generators that mimic (or sometimes steal) from a vast repository of language and art on the internet. This has resulted in a vast swathe of newly generated AI story and art for users. Users have used Chat GPT and other models creatively to help brainstorm stories, create clever commentaries, and leverage ideas more easily for consumption.

However, not all AI chat and art use is created equal. AI is most often trained on vast repositories of unlicensed copyrighted works, but provide no benefits to the underlying art and story creators. This has lead to some larger artists and authors to suing the AI companies for use of their language or work in the training process.

Generative AI also has a bad habit of creating or filtering misinformation to a broader audience. AI is becoming positively associated for its use and promulgation in neo-Nazi, supremacist, and extremist groups as a tool to create and curate false images or photos that are potentially indistinguishable from true images. It is possible for actors to create massive amounts of misinformation at the click of a button with AI.

Finally, like most precursor AI, there are concerns regarding the fairness and bias of the underlying machine application. Machine programs have a tendency to reinforce social biases within the underlying materials, in ways that become obvious in the outputs. Machines learn from human-made products in a sense, and those products are imbedded within social frameworks that still encounter racism, sexism, and bias on a regular basis.

What should society's reaction to current AI models be? What are the dangers of current, or future, AI models? Are there regulations that could help address the current issues and concerns of Chat GPT and other AI generators? What, if anything, can we do as individuals to prepare ourselves for AI generators and navigate their potential pitfalls?