Reading this made me think on the recent grants needed to sustain the AI race are honestly alarming. Every new “breakthrough” seems to come with a bigger bill attached, and the idea that only trillion-dollar companies can afford to keep up feels… off. It’s starting to look like the economics of cloud-scale AI might collapse under their own weight. On the other hand, I always looked amazed with little home clusters to run llms. Do this mean local LLMs will have their time to shine and products related to building general public tools and hardware is the big underdog here?