Toxic Truth: How Wikipedia Poisons Global Knowledge (ellakenan100.substack.com)

On its 25th anniversary, Wikipedia faces significant criticism for becoming a platform rife with bias and misinformation, impacting not only public knowledge but also AI training data. With 3.5-4 billion monthly visits, Wikipedia serves as a primary source for large language models (LLMs) like ChatGPT and Gemini. This reliance exposes AI systems to the distortion of facts, as organized interest groups manipulate content, leading to the erasure of historical narratives and the propagation of disinformation, notably concerning the Israeli-Palestinian conflict. Significantly, various communities, including the Jewish population and women, are observing systemic bias, exemplified by controversial Wikipedia entries. For instance, despite intervention from co-founder Jimmy Wales, attempts to correct biased categorizations related to Jerusalem and Israeli identity have failed, reflecting ongoing editorial suppression. This situation underscores the urgent need for fundamental changes in Wikipedia's editorial policies and a push for AI companies to reconsider their data sourcing. In response, alternative platforms like Grokpedia and new AI initiatives, such as BrightMind AI, are being developed to offer more reliable and balanced information sources, aiming to reduce dependence on Wikipedia's compromised data.