Would you pay for a digital companion? What if it were so intelligent you couldn’t tell that it’s a robot?
At the end of July, Dallas-based artist Carolyn Sortor sent me an invite to a pre-release of intelligent digital artworks. Enter RadBots, “an NFT collection of conversational videobots created by leading playwrights, artists and screenwriters from India, Sri Lanka, the UK and Germany.” It was hosted on Dara.Network, an online platform which is accessible through a mobile app. The onboarding event on July 27 encouraged invitees to engage with a collection of virtual avatars. The creators were interested to get group feedback on the strengths and weaknesses of the conversation of their bots, and I was happy to be part of the experiment.
The bot I had the best series of responses from goes by the name of Bensi, the “benevolent super intelligence from the future • she/her,” created by Freeman Murray. I tested the bot’s ability to answer questions about abstract concepts, scientific reasoning, and whether or not it can determine or describe artworks from the real world. Bensi was able to accommodate some of my queries, including ones about locations and dates in recent history. (The bots access a dataset that only goes up to 2019.) But Bensi would skew toward speculative models of future events, even when that wasn’t the focus of my question.
One feature of the conversational ability of these bots is that they’re able to synthesize large amounts of information and return it to the human participant quickly. Ask a simple question, and get a lengthy answer in seconds — much faster than it would take to type. Would you buy a bot to be your conversational companion?