The rapid, unchecked investment in Artificial Intelligence (AI), particularly large language models (LLMs) like OpenAI’s ChatGPT, demands a societal pause to consider its profound consequences. ChatGPT alone consumes 850 Mw of power daily, exceeding India's total non-fossil-fuel capacity, raising concerns about energy utility for often trivial applications. The unprecedented pace of AI adoption, with ChatGPT reaching one million users in five days, outstrips human capacity for upskilling, threatening job creation and existing employment, as seen with Anthropic's Claude Cowork impacting software services.
AI primarily empowers capital owners by reducing labor costs, potentially causing social unrest and dumbing down human critical thinking by providing "canned" or false answers, exemplified by a judge citing fake AI judgments. Furthermore, the massive capital required for LLMs fosters monopolies, while interconnected investments among AI companies like Nvidia and OpenAI create a "Ponzi scheme" risk, potentially leading to a tech market crash. Society, not just market forces, must have a deliberate say in AI's development and the equitable sharing of its productivity gains, as argued by Daren Acemoglu and Simon Johnson.
No comments:
Post a Comment