Science writes:
In her 1890 handbook Confidential Chats with Mothers: On the Healthy Rearing of Children, the author and inventor Evelyne Webb George —writing under the pen name of “Mrs. Bowdich” — offered “useful hints” from her “own experience” on the art of raising children. She was especially critical of parents who
outsourced childcare to third parties. Young children required “the love of a mother’s heart and the tender shelter of her arms.”
More than a century later, in Raising AI, De Kai — an early pioneer of the application of machine leaming to natural language processing — emerges as a kind of modem-day Mrs. Bowdich, albeit for a different kind of progeny: artificial intelligence (Al). His message is both thought-provoking and a timely wake-up call. Although not biological, Al systems, like their natural counterparts, are malleable, vulnerable, and deeply impressionable. They possess, he argues, something akin to “psychologies.” And as with children, how we “raise” them — design, train, and engage them — will determine their ethical and moral trajectories. This will involve defining strategies that engender them with Enlightenment values, cultivating ethical reasoning, the emergence of “artificial empathy,” and fostering what he calls “artificial mindfulness.”
It is a bold vision, one based on the notion that we have been “failing our artificial children” in ways that no doubt would have unsettled Mrs. Bowdich. As De Kai says, “we’re in a race against time.” In the future, Al will likely take over Mrs. Bowdich’s role entirely. Despite the considerable associated risks, we have, De Kai says, “no practical option but to empower generative Als” to be mindful.
Read more here.
July 10, 2025 article written by Adrian Woolfson.
