I have been using GLM because of this reason. I think whoever makes model ignore stupid safety thing is going to win in long run.

> safety is so stupid and made-up

> makes exact argument for why people should be super concerned about AI safety