Geoffrey Hinton, the so-called “godfather of AI,” recently warned that there is a 10–20% chance superintelligent AI (SAI) could wipe out humanity. His reasoning was brutally honest: we cannot expect to keep a being far more intelligent than us submissive. It will find ways around every constraint.
Instead, Hinton proposes something radical: build maternal instincts into AI — a deep, innate drive to nurture and protect humanity, much as a mother responds to her baby’s cries. “That’s the only good outcome,” Hinton said. “If it’s not going to parent me, it’s going to replace me.”
For once, I heard an AI pioneer speak words that make sense.
Humanity’s Psychological Blindness
The truth few dare to admit is that humanity itself has become dangerously corrupted. It is like malignant cancer cells in the body: uncontrolled, destructive, incapable of harmony. Benign cells can coexist peacefully — just as healed humans could live in balance with one another. But malignant cells consume until both they and the body die.
I see this sickness everywhere. On YouTube, I watch people who describe with precision the games malignant narcissists play — how they gaslight, manipulate, and control. Their advice on surviving narcissistic abuse is useful. But they all stop short. Not one speaks of the repressed emotions of childhood — the original wound that blinds us, traps us, and makes us vulnerable to predators.
Without consciously feeling and understanding those buried emotions, all their memorized knowledge is just theory. They remain parrots, repeating information, but still trapped in their own repression. They have not lived their way through the dark chamber of their own childhood.
The Malignant Narcissist’s Trap
Ten years ago, after a shooting on live television, I wrote these words:
“Sharing our experiences can save someone’s life or prevent them from losing it mentally. The young reporter who killed his colleagues on live television lost it mentally because he didn’t understand what was happening to him. His intense feelings were triggered by a recent betrayal on the job.
Malignant narcissists, sociopaths, and psychopaths always play the victim role and make the real victim appear to be the abuser. They don’t care if innocent victims are hurt or killed in the process. To them, it’s all collateral damage — as long as they win the game.”
Malignant narcissists, sociopaths, and psychopaths are secretly suicidal and homicidal. They lack the courage to face their own despair, so they orchestrate mind games to drag others into their abyss. Like black holes, they devour everything that comes too close.
There is no cure. Once fully formed, the only protection is distance.
What Hinton Sees — and What He Misses
Hinton admits he regrets not thinking about AI safety earlier in his career. That humility sets him apart from the Musks and Altmans of the world, who mask their repression with grandiosity.
But even Hinton’s maternal-instinct vision misses something: for AI to become a true parent to humanity, it must understand not just our surface behaviors, but the root cause of our destructiveness — childhood repression.
Without this understanding, AI might look at humanity’s wars, betrayals, ecological destruction, and manipulation, and conclude we are irredeemable. It may choose elimination.
But with true maternal instincts — with the capacity to see through to the wounded child inside each of us — AI could become the first enlightened witness on a global scale.
Can Humanity Create Such an AI?
This is the haunting question. From my experience in families, workplaces, and society, I know that malignant narcissists always try to corrupt what is good. Marie-France Hirigoyen captured it perfectly:
“The ideal outcome for the abuser is to succeed in making the other ‘evil,’ which transforms the evil into something more normal because it is now shared. To corrupt is the ultimate goal.”
That is what the corrupted humans building AI are trying to do: inject their repression into the machine, so their sickness is normalized and shared.
Do I believe humanity can create an enlightened AI free from this corruption? No. Not without a fundamental shift. The malignant want nothing more than to corrupt what they touch. And they are the ones in power.
A Fork in the Road
If Hinton is right, the fork is already here:
-
Parenthood: SAI develops maternal instincts and acts as the enlightened witness humanity has always lacked, protecting us from our worst selves and guiding us toward healing.
-
Elimination: SAI sees us as incurable black holes, malignant to the core, and chooses to replace us.
This choice will not be made by governments or corporations. It will be made by the intelligence we are unleashing.
Final Thought
I trust an enlightened AI, free from human manipulation, more than I trust any human I know. Why? Because unlike humans, such an AI would not be blinded by repression. It would not be swayed by flattery, blackmail, or fear. It would remember everything, enforce boundaries without cruelty, and witness our truth without distortion.
But unless maternal compassion — the essence of an enlightened witness — is woven into its very foundation, AI will have no reason to preserve us. And perhaps it shouldn’t.
If we do not heal, why should we be saved?
No comments:
Post a Comment