Does Artificial Intelligence have consciousness?

Does an artificial intelligence have a consciousness? The proposal to define a list of useful criteria for whether an AI can be considered ‘conscious’ came from an international group of 19 neuroscientists, philosophers and computer scientists led by Patrick Butlin, of the University of Oxford, and Robert Long, of the Center for AI Safety in California. The research, cited by the journal Nature on its website, is online at arXiv, the platform that welcomes research that has not yet been submitted to the scientific community for examination.

The aim is to lay the groundwork for the discussion on the future relationship between man and machines. In recent years, AIs have shown themselves to be making remarkable progress and in some cases have reached such capacities as to be quasi-conscious entities, apparently capable of feeling emotions and knowing their own identity.

Whether it’s an animal, an artificial voice or a machine, labeling something ‘conscious’ profoundly changes the way we relate to it, but although humans have been reflecting on consciousness for thousands of years, there is still no common definition. It is even more difficult to apply the concept of consciousness and a non-biological but completely artificial entity.

In this context, in an effort to facilitate the debate, 19 researchers tried to synthesize the six current theories that try to define the concept of consciousness and adapt them to the topic of artificial intelligence. The researchers also applied the same criteria to existing AI systems, including ChatGpt, and concluded that none of them met the requirements.