Deft, a new model focused on “better writing,” is a promising demonstration against “not X, but Y,” the patterns of three, and the words “real” and “genuinely.” DFT v1, cofounded by @jmrphy and @rosmine, entered public beta last week. The team behind Deft has a theory about why AI writing feels the way it does: LLMs don’t distribute words as naturally as humans do, with our varied sentence lengths and our range of phrases. @kplikethebird put Deft to the test by asking it to write two things:…