Did I break ChatGPT or another #Khawaja bias?

In a recent thought provoking post by Ibrahim El Badawi who reflected on ChatGPT's application towards advising on E-Government innovations, he detailed a ChatGPT bias towards calendars that begin a week on Mondays even for predominately MENA regions. Secondly, he outlined how ChatGPT's recommendations in Arabic where inferior compared to equivalent tourism recommendations in English. It is easy to often read these biases at a superficial theoretical level until they become much more visceral. This was the case for me recently when I "conversed" with ChatGPT.

My issue with ChatGPT started in a very standard and may boring chatbot like interaction where I asked it about whether it had an opinion on the best rap lyricists and whether it can provide top ten exemplars of rap flow, wordplay etc. It's answer and replies are below and for the most part, the attempt is commendable.

I took issue with a certain rap line it attributed to Notorious B.I.G rather than the rightful owner Tupac Shakur. In the process of our "conversation", I was flagged by OpenAI for possibly violating the content policy. I have already been in touch with OpenAI to discuss possible ways these sorts of conundrums could be avoided or guided or blame shared. Could this be a 2023 version of interesting ethical dilemmas that Genevieve Bell hinted at way back in 2017 during her Boyer Lectures where she suggested that in the future, your AI devices could be brought to testify against you in court.

As the image below shows, I had been questioning ChatGPT's own answer when a message from OpenAI pops up suggesting that I am the to blame for a possible violation of Open AI's content policy. Perhaps it is an over-reaction on my behalf and Open AI is enacting a form of self-censorship? In any case, however it is interpreted, it is bound to raise further interesting discussions. For example, how much more smarter and human like must ChatGPT become before it shares some of the blame for possible content policy violations? Or will ChatGPT be forever immune to blame but open to high praise for its accuracy? At the same time, can we really blame ChatGPT or other AI products for being inaccurate/biased towards subcultures within dominant cultures?

One thing that is clear is, the more voices we have on the decision making table prior to beta testing, the more GPT-3 spin-offs like ChatGPT will evolve towards futures that will benefit the sophisticated and amazing world that all of us call home.