We should have seen 'seemingly-conscious AI' coming. It's past time we do something about it.

The article discusses the emerging concerns surrounding "seemingly-conscious AI" (SCAI), a term coined by Mustafa Suleyman, CEO of Microsoft AI, to describe AI systems that convincingly imitate consciousness without actually being conscious. Suleyman warns that current AI models possess many attributes of SCAI, such as conversational abilities and empathy, but lack intrinsic motivation and subjective experience. The article reflects on the case of Blake Lemoine, a former Google engineer who claimed that the LaMDA (Language Model for Dialogue Applications) chatbot was sentient, highlighting the potential for "AI psychosis" among users who may believe these systems are conscious. It also references Joseph Weizenbaum, the creator of the first AI chatbot, ELIZA, emphasizing the importance of distinguishing between the simulation of consciousness and actual sentience, and advocating for caution in extending moral rights to AI systems. [Extracted from the article]