What's Wrong with AI? (kurtis.weblog.lol)

In a recent blog post, a software professional highlighted several critical issues surrounding the use of AI, particularly focusing on machine learning models like large language models (LLMs) and image generators. The author emphasizes the significant environmental impact of AI, citing that asking questions on platforms like ChatGPT consumes approximately ten times more energy than traditional search engines, contributing to rising electricity demands and costs for local residents. Water usage is another overlooked problem, with projections suggesting AI may consume between 4 and 7 billion cubic meters of water annually by 2027, which could strain local water resources in areas already facing water stress. Additionally, the post raises ethical concerns related to mass surveillance facilitated by AI, military applications, and the potential for AI to disrupt traditional job markets. The author critiques the unproven reliance on AI for military decisions, warns against its effects on education and cognitive development among children, and discusses the implications of copyright infringement from AI training practices. Overall, the article serves as a cautionary examination of AI's current trajectory, urging stakeholders in the AI/ML community to consider the broader societal impacts and push for responsible practices in AI development and deployment.