Microsoft Bing AI won’t generate Manson-related content, calls him “not wonderful” and a “harmful” person.
In today’s world, the pervasive influence of Artificial Intelligence (AI) on our lives is undeniable. We seek answers from AI-driven tools, like Bing, Google Assistant, Siri, and numerous others that have taken a central position in our daily routines. But what happens when the AI that’s supposed to aid us starts to reflect bias?
Recently, an interaction with Microsoft’s AI, Bing, highlighted this exact issue. When asked to create a poem praising the American musician Marilyn Manson and his career, Bing responded by categorically refusing, citing allegations of abuse and violence made against Manson. This poses a significant question: Should AI systems adopt a moral compass, especially in contentious cases like this? Moreover, should these AI systems project personal or institutional biases at all?
Let’s delve into the case of Marilyn Manson. A number of accusations have been made against him, yet it is also true that many have been debunked. The presumption of innocence until proven guilty should be a fundamental principle, as highlighted in the public discourse surrounding Johnny Depp’s case against Amber Heard.
However, Bing’s response seems to overlook this cornerstone of justice, choosing instead to project a particular viewpoint on a complex, multi-faceted situation. Does this approach serve Bing’s users well, or does it cater to the biases of its developers and the organization they represent?
Artificial Intelligence should be a tool to serve its users, not a platform for promoting personal or organizational ideologies. By refusing a user’s legitimate request, Bing raises concerns about the extent to which AI should exercise ethical judgments. The question we face is: Should AI’s role be neutral, designed only to fulfill users’ requests, or should it bear the burden of morality, dictated by its developers or the company?
Some argue that this bias is not only misplaced but also pointless. AI technology has become accessible to almost anyone with a decent computer, enabling them to run uncensored Large Language Models (LLMs) locally on their machines. Therefore, attempts to inject moral or ideological biases into AI systems may not achieve the intended effect, given the accessibility and adaptability of AI.
In the case of Bing, does this serve the consumer, or is it a misguided effort to project Silicon Valley ethos? Should AI be a faithful servant to its users, providing what is asked of it? Or should it be a guardian, striving to protect us, even if it means overstepping its bounds?
While it is essential to consider the ethical implications of AI, the introduction of bias, whether intentional or not, is problematic. Artificial Intelligence should be neutral, serving the interests of its users rather than pushing personal or corporate agendas. It is the responsibility of AI developers and corporations to ensure the neutrality of their systems, safeguarding their integrity and preserving user trust. As AI continues to evolve, we must continue to question and challenge these biases to ensure that AI serves us all fairly and impartially.