One might think a rebuttal of this article unnecessary. Yet the very fact that it was published by P&I suggests the opposite.
(Let’s be clear: I have no problem with the use of large databases to solve complex problems. Just one interesting local example is the ability of Vic Health to now predict thunderstorm asthma events with great accuracy).
But LLM generative AI is something else altogether. It sucks up all the data in the world (literally), requiring gargantuan amounts of electricity and water, while creating … what, exactly? – other than psychopathic tech bro trillionaires.
And now these authors want us to believe that “with the right training”, AI can do “relational intelligence” for us. No, no, a thousand times NO. This is AI’s biggest threat of all: that we let non-sentient machines usurp the fundamental, messy but real nature of our selves as interacting social beings. The authors should stick with but strengthen their heading: AI WILL weaken human connection – regardless of any efforts we make to “train” it.