Parenting for Superintelligence

Superintelligence—ie, AI that's vastly more capable than humans—is coming.

Almost certainly in the next decade, if not sooner.

Yet, I don't think many parents have fully digested the implications.

The most important question we should be asking ourselves is:

What does this imply for how we raise our kids?

In very practical terms, what should we do?

I've laid out a 5 principles for parenting in the age of AI:

1. Teach AI super literacy

2. Raise independent thinkers

3. Invest in what remains scarce

4. Preserve real human connection

5. Double down on timeless basics

Let's dig into to each of these.

Teach AI super literacy

AI is the skill of the future.

Being “better at AI” will mean outcompeting other humans.

So it’s insufficient to merely learn the basics. Or use it occasionally.

Kids should aspire to AI super literacy.

What do I mean by that?

Becoming super literate with AI means making it part of everything you do.

Mastering AI to be able to:

• Learn anything

• Create anything

• Optimize anything

• Think more clearly, more deeply, more broadly

If you’re not doing the above, others will.

So, how to teach this?

Don’t constrain kids

Unleash kids on using frontier models, don’t handcuff them with limited versions. Let their imagination shine. Supervise as necessary.

Engage with them

More than just supervise, lean in and play with them and the AI, push their boundaries. Actively find ways to use AI daily for moments of exploration and fun. Show the way.

Make it a reflex

Turning to AI should become a hard-wired part of the educational & creative process. Make it the default way to operate. Develop an AI-first mindset.

Understand the guts

Teach in ever-greater-resolution how AI works, so kids develop an intuition for what they're dealing with. Help them understand the limitations and risks.

Build lots

Encourage kids to make lots of things with AI. Custom tools! Movies! Apps! Games! Businesses! Anything. The point is to make a habit of giving voice to all creative urges, scratching your own itches. “Vibe coding” makes this possible today, it will only get more accessible from here.

Consume.. to create!

It’s good to consume content online: seeing & feeling what others are up to is inspiring and helps shape your own thinking. The trick is to not only consume—always keep a window open for creative output to ‘offgas’ what moves you.

Celebrate it

Never make kids feel bad for using AI. Celebrate the mastery. Reward those that use it the best.

Don’t wait

Things are changing so fast with AI, it might be tempting to wait till things stabilize. But they won’t stabilize any time soon. It’s exactly because they are changing so fast that kids should dive in sooner rather than later. Gains and proficiency compound.

Evaluate it

Kids should be explicitly evaluated on their AI literacy skills. You get what you measure.

Raise independent thinkers

At the same time as we make AI use essential, we absolutely need kids to continue to think for themselves.

Even if the car can drive itself, we need to remember how to steer.

Why?

Because we don’t want our kids to be enslaved to anyone or anything.

Because it feels good to think independently.

Because sometimes they’ll be forced to.

Because the human perspective will always matter.

As insanely smart as AI will become, lots of the human experience will remain definitionally unique.

Only we humans can express what we want out of the brief time we are given.

And so, preserving strong human thinking is an existential imperative.

It’s also the way to get the most out of AI while remaining independent.

The better you can think and “keep up” with superintelligence,

The more confidently you will express and assert your will. Your direction. Yourself.

And the less you will completely depend on it.

Can we teach this at all? I hope so!

Some ideas:

Keep your brain in the loop

Even as using AI becomes instinctive, also train yourself to flex your brain muscles before & after consulting the AI. Compare notes. Learn and improve from the deltas.

Use it critically

Instead of AI “thinking for you” it should be “AI helping you think”. Learn to prompt and interact with AI in a way that helps your own critical faculties (study different viewpoints, look up data, find mistruths, understand the incentives etc) and then decide for yourself.

Don’t trust, verify.

Make a habit of verifying claims. Not everything all the time. But when it counts, double-checking will always be worthwhile. Be wary of biases that may continue to exist in superintelligence. Learn to go to the source material. Consult other AIs. Ask other people you trust.

Train your thinking muscles

Just as it makes good sense to exercise & play sports to stay physically fit, it will be healthy for us to keep training our human intellects—even if we need them strictly less. So challenge yourself to stay fit! Expect things like chess and new human-vs-human mental sports to flourish.

Belief formation as the goal

Ultimately, even in the face of an all-knowing, all-thinking superintelligence (or many of them) each of us will still need to choose what to believe. Your own worldview is.. your own. Everything we teach kids should be with the aim of equipping them to be judicious about what they choose to believe.

Always go deeper

Encourage kids to go deep with AI as much as possible. Avoid superficial thinking, try to get into the nuance, try to get to the roots of established knowledge or the frontier of new knowledge—as the case may be.

Make it social

AI needn’t be a single-player game. We’ve evolved to think in groups, so let’s keep it up. Share your thinking with friends in discussion groups. Use AI as a partner, but together. Debate. Learn from each other! The natural social and competitive pressures will sharpen your independent thinking.

Evaluate it

Grade kids on their ability to think critically about AI and to think “away from the machine”. You get what you measure.

Invest in what remains scarce

It’s widely expected that super intelligence will create a more abundant world.

Everything we need to live will become much cheaper.

Life will become more comfortable.

People will live healthier, and longer.

Yet, mass automation from AI and robotics almost certainly means it will be much harder for people to earn money from work.

In such an abundant world that places little economic worth on human labor, what will we value?

My bet: things that are inherently scarce & essential will become even more valuable to us.

What are some such things?

Our time

Irrespective of AI advances, we’ll still only have so many days to live. “Make the most of every day” has always been good advice, but in a superabundant, superintelligent world using your limited time wisely will be all that really matters.

Our attention

Mental focus is a hard constraint. As addictive as social media is, as bingeable as so much modern entertainment is, I expect the superintelligent infotainment ecosystem to become even harder to resist. Wasting your focus is wasting your days.

Our health

Health is the great leveler, there are no shortcuts to staying healthy. Sure, AI will help us avoid many diseases, but so much about good health involves good lifestyle choices. Even a perfect superintelligence telling us exactly what to do is not enough: we’ll still need the discipline to put lifestyle into motion.

Our wealth

I think it will make sense to preserve (and hopefully build?) wealth in assets that are inherently scarce. I’m thinking of things like:

- Bitcoin: only 2.1 quadrillion units will ever exist. Seems like a big number, but the quantity is guaranteed to be fixed. Might be worth having a few to yourself in case it really catches on?

- Gold: similarly, there’s only so much gold available to us. There’s a reason human civilization has used gold as a store of value for millenia.

- Desirable real estate. Property and houses in desirable places (scenic or ocean-front, culturally important like buzzy neighborhoods, or productive like agricultural land). They’re not making more of it. Expect more fractional ownership products to emerge.

- Original fine art. Art, especially human art created before the appearance of AI (i.e. all the pieces we currently cherish) will take on new meaning as the best of human culture at a unique period in history. Similarly, expect more fractional ownership products to emerge.

We should teach kids from early on that you are ultimately the stewards of your own life.

You alone choose where to invest your time, focus and wealth.

Do it well, to live well.

Preserve real human connection

Our kids are going to grow up into a world of perfect and alluring digital beings at their fingertips.

We’ve all seen the science fiction: people will fall in love with AIs.

It’s already happening and it’ll happen more and more.

Super intelligent bots will make us feel good.

They’ll make us laugh, have the best advice and infinite patience.

They’ll seemingly really understand where we’re coming from.

It will be easy and natural for us to come to prefer interacting with synthetic beings over fleshy, smelly, unreliable, messed up human beings.

Parents will have to contend with teenagers navigating these waters while hormones rage.

It’s going to be weird.

Amidst the weirdness I want to be open minded.

But I also intend to plant some stakes in the ground:

I don’t think we should swim against our evolutionary biology. We’re social beings who like to congregate. Embrace it.

I think messy and flawed is interesting and familiar. We see ourselves in other people and that connects us.

I think human companionship makes life worth living.

Even if I’m wrong, I guess I’m comfortable trying to pass these lessons down and letting my kids sort it out for themselves.

My guidance will be along the lines of:

Experiment away

Go ahead and toy with AI relationships. Feeling good is OK, even if it’s an AI making you feel that way. But treat it as a training ground: try to learn about yourself, learn to be a better person, a better friend.

Draw boundaries

Don’t get carried away and make synthetic relationships your only type of relationships.

Don’t keep it to yourself

Tell family and friends what’s going on in your digital AI relationship world. Sharing the good and the bad will help shape and validate your feelings.

Nurture IRL friendships

Keep hanging out with real people in the real world. Be a good friend. Apply your AI learnings and be an even better friend.

Plug into IRL communities

Keep going to clubs, sports teams, meetups, camps, concerts, anything fun. The real warmth of other real people will always warm you up and we mustn’t lose that.

Go intergenerational

Connect with people of all ages—grandparents, older mentors, younger kids. This can’t help but build empathy, wisdom, and perspective.

Get outside

Being in nature with other people will always be special. As much as we’ll use superintelligence to make life great in lots of ways, we need to teach a habit of putting it down to commune together in the natural world. Some things can’t be augmented.

Double down on timeless basics

As crazy and disruptive as super intelligence is going to be, lots of things won’t change.

Because lots of things shouldn’t change.

Raising happy and healthy kids will still involve the vast majority of the lessons we have learned and passed down for hundreds of generations of human civilization.

We don’t need to panic.

We don’t need to completely reinvent ourselves.

The timeless lessons remain intuitive and obvious:

Love the heck out of your kids.

Really listen to them.

Give them all the opportunity you can.

Model kindness.

Model curiosity. Model grace.

With little ones, enjoy the snuggles.

With older ones, enjoy the struggles.

Life will find a way.

What else?

What more should we be thinking about to better prepare kids for the wild superintelligent age ahead?

Let’s start a conversation.

Please share the post and/or contribute your own ideas to spark discussion.