Many people now think training a model is writing a large markdown doc…

→ View original post on X — @jiquanngiam