I'm not sure about this story on computers grading standardized tests, because I don't know about the sophistication of the software or the levels of the students. My gut is to agree with Tim Oates, director of Cambridge Assessment:"Some approaches look like technology in search of a test, rather than assessment designed to accurately report attainment."
We've started a rather successful rhet comp reading group this semester--our first two readings have dealt with the "image" of teaching grammar and "correctness." This push toward automated software will exacerbate the problem rhet/comp has explaining why teaching grammar is often antithetical to teaching writing. Grammar is a matter of a-contextual order and precision, writing is a matter of highly contextualized and nuanced choices and experiment. That's why the medieval trivium included bothgrammar and rhetoric. And medieval universities assumed that a student needed four years of both subjects (along with logic--hence the 3) to gain proficiency.
Mistake Two: Terence Kealy provides an attempt at humor gone horribly wrong. "Let male professors sexually fantasize about their female students." Yeeeesh. Might as well make a monkey joke. Oops, that's on us.