AI Consciousness is a Big Deal. Blake Lemoine famously left Google in 2022 because he thought that LaMDA was conscious. Anthropic’s Claude has a terrible habit of claiming to be conscious in long conversations. Science fiction is full of stories of conscious AIs being enslaved and mistreated by humans. Is that what we’re doing right now??
In 2023, Butlin and Long surveyed existing work on consciousness studies and defined indicator properties using different schools of thought on what it means to be conscious: recurrent processing theory, global workspace theory, computational higher-order theories, attention schema theory, and predictive processing theory. These indicator properties serve as a rubric for assessing the likelihood of consciousness in AI systems. Systems that exhibit more of these features are considered better candidates for consciousness. Evaluating multiple systems against these indicator properties, including GPT-style transformers and DeepMind’s embodied agents, they find that no existing AI system is a strong candidate for consciousness.
Let’s read Butlin and Long, bone up on the science of consciousness, and then ask ourselves whether “consciousness” was the thing we should have been worried about in the first place.
Whether current or near-term AI systems could be conscious is a topic of scientific interest and increasing public concern. This report argues for, and exemplifies, a rigorous and empirically grounded approach to AI consciousness: assessing existing AI systems in detail, in light of our best-supported neuroscientific theories of consciousness. We survey several prominent scientific theories of consciousness, including recurrent processing theory, global workspace theory, higher-order theories, predictive processing, and attention schema theory. From these theories we derive "indicator properties" of consciousness, elucidated in computational terms that allow us to assess AI systems for these properties. We use these indicator properties to assess several recent AI systems, and we discuss how future systems might implement them. Our analysis suggests that no current AI systems are conscious, but also suggests that there are no obvious technical barriers to building AI systems which satisfy these indicators.