I start from CAPTCHA, those distorted strings of letters set against colors and designs, which we all recognize and rewrite to gain access to web sites. CAPTCHA operate in a complex intersection between commercial and scientific institutions, on the one hand, and concepts of mind and being, on the other. The Turing test is the background supposition of CAPTCHA's claim that recognition and rewriting is the iterative presentation and production of a singular human self. It is iterative and weak, in that it must be repeated and rewritten, and in that it is open to failure and misrecognition. It is singular and productive in that each test ties to an event that can be leveraged commercially - as in the reCAPTCHA project or (I argue) in related "distributed work" programs such as the Amazon Mechanical Turk - and in that the "weak performativity" couples the test to a range of net activities on the part of a speaking subject. (I contrast this to the traditional linguistic "shifters," and instead approach the subject in terms of a Kleinian "psychoanalytic graphology.")
(Source: Author's abstract for ELO_AI)