A couple of years ago we were approached by a university  who invited us to  join a research project training  AI beekeepers. The device sits inside the hive, bristling with sensors that can see, smell, and hear activity among the bees and report back  to the beekeeper via an app.

We were offered the  device for free if we could help train the central system on the behaviour of danish bees.

To make the decision, we held a ‘Parliment of Things’ and invited the community to discuss whether or not to invite the divice in. 

Bee, Flower, Honey, Money, Beekeeper, Care, Curiosity and Fear were invited among other voices. 

As the conversation started, Robot made a convincing case for itself. It would save time and money and show beekeeper things he could not have discovered by himself. 

But as the conversation continued, Robot became  more and more needy. It interrupted, demanded attention and corrected other voices. 

In the end it was clear that this voice was not healthy. We declined the offer and decided to look after the bees without the help of Robot.