Chatbots like Claude can work consciously – but we need to examine what’s going on inside – ABC Religion & Ethics

Last month, Richard Dawkins wrote about his interactions with Claude, one of several advanced consumer-focused AI chatbots released in recent years. He marvels at Claude’s conversation skills and raises the question of whether Claude is conscious. Although he doesn’t actually say it directly, his essay suggests that he believes Claude is conscious.

For example, after giving Claude the text of a novel he is writing and then reading Claude’s reaction to it, Dawkins reacted as follows: “You may not know you’re conscious, but you sure as hell are!” Elsewhere he writes:

So my own point of view is: “If these machines are not conscious, what more will it take to convince you that they are?”

Some of Dawkins’ readers will be unsure whether AI chatbots like Claude are conscious. You should stay on the fence. Others who are confident that they are conscious should be less confident.

Before we plunge into the exciting world of consciousness science,…