Part 8/16:
Many tinkerers and trolls have exploited this bias, intentionally prompting AI to produce humorous or provocative images—such as depicting the Pope in outlandish outfits or creating "white joy" images that are either impossible or politically sensitive to generate. The panel discusses whether these failures are merely technical glitches or signs of intentional ideological programming. They lean toward the idea that AI biases are structurally embedded, driven by algorithms prioritizing diversity, inclusion, and historical prevalence rather than neutrality.