The AI History That Explains Fears of a Bubble (time.com)

Concerns are rising among investors regarding the sustainability of the AI sector, particularly as it has played a crucial role in preventing an economic downturn. Nvidia, the leading supplier of AI chips, has surpassed a $5 trillion valuation, while OpenAI, the creator of ChatGPT, continues to operate at a loss, consuming billions annually. Despite these financial concerns, venture capitalists invest heavily in AI startups, driven by the belief that AI will revolutionize various industries and transform the job market. However, the AI community faces a significant evaluation challenge. While advancements in large language models (LLMs) show promise, traditional metrics and benchmark tests are inadequate for assessing their effectiveness in complex, creative tasks—unlike the more straightforward problems tackled by earlier AI technologies. This gap in reliable evaluation methods raises the specter of repeating historical AI cycles of hype followed by disappointment. Without robust systems to measure progress accurately, there's uncertainty about whether the current AI boom represents genuine innovation or indicates the formation of another unsustainable bubble.