E-mail based generative AI: a good idea badly executed

If you set aside the gendering of the name, as well as the obvious ambitions to eliminate remote assistants (whose salary and conditions have already been crushed by freelancing platforms) there’s a narrow sense in which this could be an effective proposition taken up by a lot of people. But at present Mindy is producing GPT 3.5 levels of response which are pretty useless for the real world tasks I tried with the service. Watch this space though, if they get this working it could be commercially successful.