The Download: how to run an LLM, and a history of “three-parent babies”
In the early days of large language models, there was a high barrier to entry: it used to be impossible to run anything useful on your own computer without investing in pricey GPUs. But…



















