From this Essay, Jonathan Birch went on to have success with his book The Edge of Sentience: Risk and Precaution in Humans, Other Animals, and AI (2024).
‘Intelligence and sentience are not the same. Intelligent systems need not be sentient, and sentient systems need not be intelligent. I first got interested in the idea of artificial sentience through the OpenWorm project, an attempt to emulate the entire nervous system of a tiny worm, neuron by neuron. I thought: if we can do that, we will soon move on to emulating fruit flies, bees, fishes. Could an emulation of the brain of a sentient being itself be sentient? If so, it might achieve artificial sentience without much intelligence. Years later, we have remarkable chatbots, but they are nothing like OpenWorm. They excel at role-playing and mimicry, and so they’re able to game some of our criteria for sentience. In our essay, Kristin Andrews and I say: to find real criteria for sentience that go beyond mimicry and role-playing, we first need to understand sentience in animals.’
– Jonathan Birch
|