Erring to be Human

Few days back when I caught up with a cousin of mine, who works in one of those big consulting firms, he told something that I found really interesting (and amusing too). He said his team uses GenAI extensively for proposals, presentations and client communications. It is faster and better. Clean language, proper structure, and….perfect.  But he has a curious advice to his team, before the final version goes out, he asks his team to leave small mistakes in the document - an awkward sentence, or few typos, a wee bit of inconsistency, etc. just so that the person receiving the document to feel the ‘human effort’. This intriguing behaviour the team is showing…using a sophisticated technology to make their work perfect and then deliberately putting a little imperfection back into it so that the perfection does not become suspicious is what made me curious to why are we deliberately erring to be human.

We humans are culturally wired for not being perfect. In many beliefs and faiths, reaching perfection, is a lifelong human endeavour..and in some instances - multi-lives.  Be it the Zen tradition of Ensō - an effort to draw a perfect circle in a single stroke, or the idea of ‘warding off the evil eye’ in many traditions. The nazar-battu, the kala-teeka, evil-eye hangings, a pumpkin or even an ugly-looking mask hung on the façade of a building under construction, these are our ways of avoiding the ‘too-good-to-be-true’ syndrome and to stick to our collective imagination that human efforts are not perfect.

In psychology, there is something called the pratfall effect. When someone who is perceived as highly competent, a small and harmless mistake could sometimes make that person more likeable. The mistake seemed to make an otherwise impressive person feel more human. Such fallibilities reminds that competence is about what one can do, while their imperfections remind us about a ‘person’ who is doing it. 

Maybe, for humans, being loveable has always mattered more than being perfect.

AI can make the sentence smoother, the presentation more coherent and the photograph more beautiful. It can remove hesitation from our writing and inconsistency from our arguments. The reason we are putting the effort (for imperfection) is to show that someone was thinking rather than simply producing. The imperfection becomes valuable precisely because it is evidence of the person who made the thing. We have started to look at imperfections not as things to correct but as clues to authorship. A slightly awkward sentence may reassure us that somebody actually wrote it. An unexpected phrase may make an email feel more personal. A tiny inconsistency in a presentation may tell us that someone was making choices rather than simply asking a system to generate the finished answer.

Perhaps this is one of the stranger consequences of AI. It may not merely change the way we work; it may change what we value in the work. When competence becomes abundant, the small signs of human effort may become scarce. And when something becomes scarce, we begin to notice it differently. The typo that once embarrassed us may become reassuring. The uneven line may become a signature. The sentence that could have been polished a little more may actually make us trust the person who wrote it a little more.

We have spent a very long time trying to get the human error out of everything we make. It would be rather ironic if, just as machines become extraordinarily good at doing that for us, we begin putting a little of it back. Not because we have forgotten how to make things perfect, but because sometimes the imperfect thing tells us that someone was there to make it.