I have known Andrew Pery for a long time (I call him Andy) and have always been impressed. When I first met him more than 45 years ago, he was a recent law graduate working for legal aid. His moral compass was evident back then and has continued to this day in both work and life.
Andy calls himself an ethics evangelist. Ethics evangelists promote and advocate for what is morally right and just, and encourage transparency, integrity, and truthfulness in business. Currently, Andy is the AI Ethics Evangelist for ABBYY, a global leader in Intelligent Document Processing and Process Discovery and Mining.
On LinkedIn, Andy posted comments about an interview at the Ai4 2025 Conference in Las Vegas involving Geoffrey Hinton, the father of neural networks. During it, Hinton stated, “We need AI mothers rather than AI assistants. An assistant is someone you can fire. You can’t fire your mother, thankfully.” Hinton believes we have to prepare ourselves for machines that will outthink us and that they will be here within a decade. Maternal instincts may be the only way to get AI committed to our survival, since the money being spent by Big Tech in the race to be the first to superintelligence seems to have no stopping point unless regulators create guardrails.
I’ll let Andy describe the challenge that lies ahead.
Geoff Hinton’s latest warning is that our survival may hinge on building AI with “motherly instincts.” That’s a provocative metaphor, but the impediment isn’t technical imagination. It’s incentives. As long as engagement and speed-to-market outrun safety, “ethical AI” will remain a slide, not a strategy.
Look at the political economy of today’s platforms: our public sphere is funded by a grand bargain, free services in exchange for pervasive data extraction. That market design rewards attention capture and opacity, not care and accountability. Asking these firms to prioritize safety over profit without changing the rules is like asking a casino to lose money for the vibes.
So, ditch the lullabies. Regulate the incentives.
Jack Balkin of Yale Law School, in homage to the first creator of laws for AI and robotics, Isaac Asimov, has created the Three Laws of Robotics in the Age of Big Data. They are:
- Big tech must be treated as information fiduciaries: Â Treat dominant AI/platform companies as owing duties of loyalty and care to their users. No exploiting data against user interests; structure governance around trust, not just consent flows.
- Make companies internalize the social costs of generative models: don’t let “externalities at scale” be a business model. “In an algorithmic society, injured parties cannot easily absent themselves from seeking jobs, housing, and medical care,” where algorithms will determine their quality of life in a digital world.
- Impose objective liability standards: When firms deploy “risky agents,” the law should apply a strict liability standard with a presumption of causality imposed on providers of AI systems, given their asymmetrical control of AI system design.
If Big Tech wants the keys to cognition, it must accept the duties that come with power, be a fiduciary, prevent algorithmic harms, and impose strict liability when things break. Anything less is just branding.
Can we honestly expect that Big Tech would be incentivized to embed “motherly instincts” in AI?  Quite the contrary.
The AI race is now in full frontal mode. It’s all about AI dominance. As the maxim goes, “Winning is not everything. It’s the only thing.”
