The development of forward mode automatic differentiation rendered decades of computational statistical work on numerical approximations through linearizations obsolete.

Le Cam wrote his likelihood papers and books before the end of the 1950s, computational statistics spent decades finding numerical approximations through linearizations for increasingly complex models, only for for forward mode autodiff to make the computational work irrelevant

Le Cam wrote his likelihood papers and books before the end of the 1950s, computational statistics spent decades finding numerical approximations through linearizations for increasingly complex models, only for for forward mode autodiff to make the computational work irrelevant