AI (Sort of)

This week, we spent a lot of time talking about AI, and machine-learning algorithms vs rules-based algorithms. For example, our fortune teller could be mistaken for an AI (if it was a bit more refined, of course) as it works similarly to a chatbot like ChatGPT. This was the distinction that we made this week: rules-based programming isn’t “artificial intelligence” because it isn’t learning and adapting from the conversation.

This is exactly how ELIZA works! We took some time to chat with ELIZA, an early “chatbot”, who really likes to keep the conversation going. Dr. Whalen dove into ELIZA’s source code, and we were able to tell how she worked. ELIZA is parsing the prompt that we, the user, input and is reforming it into a question that keep conversation flowing. It feels almost exactly like a therapy session, or small talk with a friend who really wants to know how you are doing.

We also touched on popular representations of “AI” in media; we had to select a fictional artificial intelligence and share them with the class. Anna and I chose GLaDOS from Portal, but the very first one that popped into my head was Jarvis from Marvel. I have been seeing a lot of videos, both on YouTube and TikTok of hobbyists recreating these AIs; one TikTok creator has made a Jarvis-like interface in his workshop that aids in his 3D printing endeavors, but also works similarly to something like an Amazon Echo or Google Home.

Moving into the Mcllwain reading, we learned about ALERT II, a criminal justice information system. It was supposed to act as a means for identifying criminals based on prewritten algorithms. However, the “sinister shit” that lays “buried between lines and lines of technical jargon” is referring to the inherent biases that are written into an algorithm such as this. In this course, but especially in Dr. Rao’s ChatGPT and Generative AI course last semester, we discussed how much bias can bleed through an algorithm, AI or not. Racism and prejudice is very much so around today, and seeing it leak into this next wave of technology is terrifying, seeing as this is one of the predominant generative forms of technology today.