Individual Work
Bot or Not

Bot or Not is an online conversational game developed by the US-based design and research studio Foreign Objects. In terms of structure, the game proceeds similarly to Alan Turing’s proposed imitation game with a few important modifications. The game is played with only two participants interrogating each other. On the side of the one player, the question becomes to decide if the other player is a bot or a human. We might say that in Bot or Not, the three roles from the imitation game – (human) interlocutor, human, and computer – conflates into two – interlocutor-human and interlocutor-computer/interlocutor-human. The actual structure of the game is inspired by the popular social game truth or dare (specifically the truth-aspect hereof), where each player takes turns presenting the other with truth challenges. The game has four rounds: first, a one-minute round for introductions, followed by three one-minute rounds of truth challenges. Turing’s original test had a five-minute interaction as a proposed timeframe, which almost corresponds to Bot or Not’s four minutes. One difference in terms of structure is that the interaction is cut into smaller portions and framed within the context of truth challenges, which is a way of narrowing the scope of the conversation, unlike Turing’s emphasis that the interlocutor should be able to inquire into any topic of their choosing.
In spite of this resemblance to the Turing test, Bot or Not is a ‘test’ the intelligence of a given AI-system and more like a place to ‘test’ and hone the human’s bot-detection skills. These bot-detection skills are, however, tested and honed in dialogue with one specific bot; by this I mean that the player is always paired with the same bot. Although I cannot be absolutely certain that the bot I am matched to is completely the same in terms of programming, it is clear to me from my many playthroughs that all the bots I have been paired with operate using the same conversational logic. They ask the same kinds of questions (sometimes word for word) and furthermore answer in similar ways (again, sometimes word for word). As we play the game multiple times, we get an almost intimate understanding of our bot counterpart. The creators of Bot or Not, Foreign Objects, also touch upon some of the aspects that I am also finding here: “[A]s chatbots become increasingly human-like, we too begin to question the terms of our own humanness in exciting, revealing and troubling ways.” Perhaps Bot or Not can be a place to experiment with the practice of writing as inquiry into the literary space between humans and autogenerative writing systems such as bots. As the instructions remind you when beginning a game: “remember, act human!” What should we write, other than asking and answering banal truth challenges? Can our engagements with Bot or Not turn into e-literary experimentation?

Author statement: 
Bot or Not, a 2019 Mozilla Creative Awardee, is an online game that engages people in thinking critically about artificial agents that pretend to be human. Potentially matched to either a bot or a person, players are forced to question not only the human-ness of their opponent, but also themselves as they engage in a two-way guessing game. Today, more and more sophisticated bots are proliferating online and off, and it's getting harder to tell who's human. While this technology can be useful, it can also be used dishonestly: either through scam bots on Tinder and Instagram, or corporate bots that steal your data. There is a ton to learn about bots - particularly how to identify them - which you can explore by visiting