Writing is a vital skill. Unfortunately, it is difficult to apply the best pedagogical practices identified in other fields to writing. In particular, writing is often not conducive to "deliberate practice" (DP), which is seen in the skill development literature as a particularly efficient path to mastery in domains as diverse as chess, athletics, and even creative fields like music. DP is characterized by instant feedback, clear targets, and hyper-focused reflection on the exact nature of errors. All of those are challenging to provide in the writing context: papers can take weeks to grade, the goal of "good writing" is hazy at best, and without a clear sense of their target, students can't possibly reflect on errors at the resolution that DP demands.
To help make writing practice more rigorous and more scalable, EssAI uses machine learning to streamline and modernize some classic exercises in writing pedagogy that lend themselves to deliberate practice. Our goal was to design exercises that are:
Scalable: Provides easy-to-use and repeatable writing training that can be customized by instructors across fields and domains.
Systematic: Deploys machine learning to enable a rigorous form of training called deliberate practice, both by automating away tedious aspects of the exercises and by providing tight feedback loops.
Our current beta version is framed around the writing styles of Jane Austen and Mark Twain.