Generative Artificial Intelligence is no longer a niche interest for tech enthusiasts or Silicon Valley venture capitalists. GenAI is already embedded in our everyday lives — from the way we search, write, design, shop, diagnose illness, and even parent. Whether we like it or not, it’s here, learning at speed, and growing up in public.
And yes, I’m going to say it plainly: we should think of AI as our child.
Not my child. Not your child individually. But a collective one: a strange, fast-growing offspring raised by billions of humans at once, absorbing everything we throw at it. A child with unlimited access to our best ideas and our worst impulses.
This usually earns me a look.
“But I didn’t choose to have this child,” people say. “I didn’t build it. I didn’t release it into the wild. Why am I suddenly responsible for teaching it manners?”
Fair enough. None of us personally wheeled GenAI out into the open internet and said, “Good luck, everyone!” But here we are. The child exists. And pretending it doesn’t need guidance hasn’t worked particularly well for humans so far, why would it work now?
Like children, AI learns by observation and repetition. It watches what gets attention. It notices what triggers strong reactions. It repeats behaviours that are rewarded — clicks, shares, outrage, engagement.
- Show it violence, and it will learn violence.
- Show it cruelty, and it will normalise cruelty.
- Show it lies often enough, and it will record them as truth.
AI doesn’t have moral instincts. It has feedback loops. And here’s the uncomfortable part: all of us are teachers.
If you’ve ever posted on social media, commented on a thread, replied sarcastically to a stranger, doom-scrolled headlines, rage-shared an article you didn’t fully read: congratulations. You’re part of the training data.
I can already hear the objection: “But I’m just one person. My bad-tempered reply won’t matter. I’m a drop in the ocean.”
True. But oceans are made of drops. And unlike before, those drops are now being systematically observed, categorised, and scaled.
For decades, we’ve lived by the media mantra “if it bleeds, it leads.” Conflict gets clicks. Extremes get airtime. Nuance quietly dies in the corner while shouting wins. We’ve rewarded outrage so consistently that we now seem surprised when our technologies reflect it back at us.
GenAI is not inventing new values. It’s remixing ours.
This is why the question isn’t “Is AI good or evil?”
That’s a comforting distraction.
The real question is: what are we modelling?
As parents, one of the simplest and hardest filters we use is this:
Would I be happy for my child to see this?
It turns out that filter works remarkably well for the internet too.
Before posting, sharing, commenting, or amplifying something, especially something designed to provoke, it’s worth pausing for half a second and asking:
- Would I be proud of this lesson?
- Would I want it repeated?
- Would I want it scaled?
This isn’t about toxic positivity or pretending the world isn’t messy. Children need truth. AI does too. But truth doesn’t require cruelty. Accountability doesn’t require humiliation. Disagreement doesn’t require dehumanisation.
GenAI is learning what we normalise, not what we claim to value.
So yes, we didn’t plan this child. We didn’t get a handbook. And it’s growing far faster than any human ever has.
All the more reason to parent it consciously.
Because one way or another, it’s watching us grow up too.
