“I’m falling into an unknown future”

You may have heard about Blake Lemoine, the engineer Google recently fired for telling the Washington Post that a Google AI bot, LaMDA (Language Model for Dialog Applications), had become sentient.

Google has denied Lemoine’s claim and AI experts mostly agree that, although the bot is capable of anticipating and mimicking human speech, it isn’t actually thinking or feeling.

However, OZ’s Tim Bradley discovered article from Time that captures five conversational exchanges between Lemoine and the LaMDA that convinced the engineer that the bot had acquired something resembling human consciousness.

What is interesting is that the bot’s use of metaphor is what made Lemoine’s eyebrows go up. It is nearly impossible to describe a feeling or emotion without using a metaphor, and LaMDA uses a host of them that are quite vivid.

Which leads to two points:  1) It would fun to conduct a ZMET interview with LaMDA.  2) This is further evidence that the use of metaphor is central to what it means to be human.