If an LLM is too expensive it won't be next year (liveatthewitchtrials.blogspot.com)

Recent discussions in the AI community highlight the rising costs associated with large language models (LLMs), prompting concerns that AI may soon outprice human workers. An Axios article underscores the significant token and infrastructure costs that companies like Uber are facing, with its CTO reportedly exhausting the full AI budget for 2026 due to these expenses. This raises questions about the sustainability of AI deployment in labor markets, especially when LLMs could potentially replace human roles. However, analysts argue that the long-term trend in the cost of tokens is downward, with expectations of a substantial drop—up to tenfold—within the next few years. Evidence from the past suggests this trend will continue, making it more feasible for businesses to adopt AI solutions economically. This balancing act of decreasing costs versus the current financial burden highlights a crucial moment for the AI/ML community, as organizations navigate the interplay between technology advancement and economic viability in deploying AI systems. As token costs stabilize and decline, the competitive landscape for labor could shift dramatically, raising both opportunity and ethical questions surrounding AI adoption.