"Scaling LLMs to get us to AGI like we scaled internal combustion engines to reach lightspeed." - @halvarflake