I think the next step would be predictive text. Have 3+ symbols that correspond to a screen that is in the subject's eye line. Basically just leverage the current tech to streamline this.