Toward the beginning of the 2002 film, a domineering lawyer (played by James Spader) barges into the office of his assistant (Maggie Gyllenhaal) with evidence of a work infraction: a memo she has written that has "three typing errors." "Do you know what this makes me look like to the people who receive these letters?" Setting aside that his screed turns out to be foreplay, Spader's character was channeling a widespread cultural revulsion: Typos were the ultimate shorthand for careless work. A spelling mistake was proof that the writer hadn't bothered putting much effort into a piece of correspondence, that their instructions or advice shouldn't be taken seriously--and perhaps that the recipient shouldn't invest time in reading their note at all. More than two decades later, as AI-generated writing has flooded workplaces, social media, and dating apps, old hallmarks of sloppiness--typos chief among them--are getting a new gloss. Some job applicants are intentionally adding typos to their cover letters to prove that they, and not an AI program, wrote them.

Full (exact) conformal set vs. split or cross-validated conformal set Non-connectedness of the conformal prediction set. This was initially suggested in [18, Remark 1]. We follow the actual practice in the literature [14, Remark 5]. We did not observe violations. We will also summarize the proposed algorithm in a direct pseudo-code.

We will fix them in the final version.