Accelerating AI development and improving safety are inherently contradictory. It's pretty annoying and disingenuous when someone says "this move will speed us up and make us safer".