Let me tell you about my mother

On the matter of bots, tortoises, and deserts in the hot, baking sun, I think social UX or social interaction design might have a second life (!) in the design of bot and AI-based interactions. At this point it doesn’t take a blade runner to catch an errant bit of AI, but in all the discussion I’ve seen on machine learning and chat bots, the focus has been on the task of acquiring languages, building domain knowledge, and “understanding” human expression. And in the shift from solving for search (by means of answers) to performing actions (doing what the user asks), machines are having to learn human commands, or “intents.” Obviously, the more our macbines know about our intents, the better they can serve us, right?

But what of the machine’s intents? They may begin with correct interpretation of the user’s requests and commands. But surely they hide their own agendas, too. Marketing, sales, promotions — the ad sense of the future will monetize natural speech. And as surely as the machines learn our intents, we will discern theirs as well. And dissimulate and mislead the machines, we will, if only to randomize our customer profiles and confound the algorithms.

It’ll be interesting to see where social interaction design fits into this, as the design of bots develops the nuance for interactions, and as transactional conversation (and not just instrumental conversation) begins to approximate not just natural language but natural interaction. We will, I’m sure, find ourselves asking our bots about the tortoise in the desert.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.