Gemini AI … Racism or Glitch?

Despite the controversies and negative press surrounding Google's latest AI model, "Gemini," I believe it has made significant contributions to humanity, albeit unintentionally.

?Gemini, Google's newest AI tool, produced images that were not accurate and sometimes conveyed inappropriate meanings, such as depicting American founding fathers as black individuals or a female pope, leading to accusations of "racist and anti-civilizational" and being "woke" by figures like Elon Musk and publications like the New York Post.

?What Gemini demonstrated is a well-known challenge in the AI domain: bias. Despite the numerous challenges AI faces, bias remains one of the most difficult to identify and rectify, as it is deeply ingrained within AI systems. The issues with Gemini could have arisen from various factors, including the data fed into the model, the algorithms used, or even intentional actions by the developers.

?Fortunately, Google has ceased Gemini's ability to generate human images due to its propensity for inaccuracies. Google's senior vice president, Prabhakar Raghavan, acknowledged that while they cannot guarantee Gemini won't occasionally produce offensive results, they are committed to addressing any issues that arise.

?Now, it falls on Google's shoulders to rectify this issue. However, what about our responsibility? While we were able to identify bias in Gemini's image generation, what about textual biases? How significant are they, and who can evaluate and validate them? The task of evaluating and certifying the results of advanced AI models is challenging, given the vast amount of data they produce and the cultural differences between countries.

?Currently, each country has its own quality control agencies, such as the FDA, EMA, CDRH or OSHA, which monitor products consumed by citizens. Similarly, there is a need to extend such oversight to the information generated by AI models, ensuring it aligns with cultural sensitivities and values. While this task is daunting, it is essential to begin addressing it.

?Until then, whenever AI models like Gemini, ChatGPT, or Lex provide information, it is crucial to ask for references and critically evaluate them.