Continental Divides

Living Only Gets Harder

Without built-in safety measures the same artificial intelligence that is helping kids with their homework today could eventually achieve “superintelligence”

By John McCaslin

I’ve long believed it will be the women of this world who prevent mankind’s extinction through whatever disastrous mechanism of his choosing.

I reached the morbid conclusion having witnessed my mother’s ability to be strong and decisive while simultaneously compassionate and caring. Most every other mom in the neighborhood followed suit.

Growing up a baby boomer during the Cold War years the risk of nuclear war topped all doomsday scenario lists. I remember rushing to my mother’s arms every time the practice air-raid siren sounded.

The post-9/11 age of “modern terrorism” would usher in more localized threats to humanity: engineered chemical and biological agents to radioactive dirty bombs that fit into a backpack.

Society’s near-total reliance on computers, meanwhile, has fostered new concerns of catastrophic cyberattacks and global system breakdowns of infrastructure and supply chains: power grids, telecommunications, transportation, food and water supplies, healthcare, financial institutions, federal, state and local governments.

Then five years ago, when Covid-19 reared its ugly head, it was runaway pandemics that kept me up at night.

If those potential calamities aren’t enough to lose sleep over (remember, I’m skipping the natural causes of carnage, from climate change and killer asteroids to supervolcano eruptions reminiscent of earth’s previous “Great Dying”) we’ve now entered the unpredictable if not precarious era of artificial intelligence.

And conceivably on its heels, God forbid, “runaway AI.”

Without built-in safety measures, which currently don’t exist, the same artificial intelligence that is helping kids with their homework today could eventually achieve “superintelligence,” surpassing human intellectual capability and escaping man’s control.

Picture AI on steroids: more clever and powerful than mankind and seeking its demise. Unless our mothers—artificially speaking—come to the rescue first.

“I wish I’d thought about safety issues, too,” the Nobel Prize-winning “godfather of AI,” Geoff Hinton, admitted at the recent Ai4 industry conference in Las Vegas.

Before losing control of our destiny, the former Google executive is recommending most urgently that “maternal instincts” be built into rapidly expanding AI computer systems.

Compassionate versions “that really care about people,” Hinton explained. “We need AI mothers … that’s the only good outcome.” 

The question is how to effectively create these life-saving instincts and strategically insert them into AI’s already advanced realm.

And could the nurturing intuitions be fashioned after our mothers as a whole or might the maternal instincts of individual women be cloned and cybernated? 

Either way, I hereby propose excluding this country’s current slate of female leaders (you know which ones) whose combative and harsh demeanors would only further infuriate the conniving AI overlords and hasten the amount of time until we’re toast.

There does happen to be one woman in particular though—her legacy the very preservation of life—who could prove successful in keeping the AI villains at bay. 

I’m referring to the late Jeannette Rankin, Montana Republican.

The first woman elected to Congress, Rankin was the only lawmaker among hundreds on Capitol Hill to vote against entry into both world wars.

“As a woman I can’t go to war, and I refuse to send anyone else,” she stood up to proclaim, “for war is the slaughter of human beings … on as large a scale as possible.”

Which is precisely what she’d be preventing.

As Hinton pointed out: “Super-intelligent caring AI mothers … don’t want us to die.”

John McCaslin is a longtime journalist and author who lives in Bigfork.