What should you teach your kids now – and in the future?
TLDR – three skills that age well:
Architect– systems thinking, creativityManager– bullshit detection, execution, coordinationDiplomat– EQ, negotiation, relationships, knowing what you want, ethics
Considering how everyone got their asses on fire with full automation acceleration now.
I think it follows exactly the same pattern we have in software development.
A localized area of life transformation, yet universal? Just a matter of time before anything non-decision-making gets an interface and some ML app to use it.
I think you can extrapolate quite easily, with high confidence, to the entire society, any kind of job.
Whatever kind of future it would be.
I am not talking here about jobs being lost and so on, just about the nature of the work itself.
I think it’s also timeless – even if all bubbles pop, and it all crashes down – those principles are kinda universal.
Would be relevant as long as humanity exists, and even if not – still!
Except for real frontier science – most of the work would be – managerial-like.
Not heavy/repetitive physical or intellectual labor.
But managing others onto that.
So, what should you, or anyone, start learning?
Well, first accept that learning is constant and continuous, if anyone still denies that.
I would name all three main areas of skill:
Architect, Managing, Diplomacy.
Three words. They look different, but at the bottom it’s the same thing – making decisions. What to build, what to delegate, what to negotiate, what to even want. Each skill is a different lens on that one act.
An architect of everything.
Let’s start with the first one, maybe the most important, as the word does not reflect the wideness of it.
What being an architect means – being able to understand a system as a whole, or the fancy term "systems thinking".
Applied to everything! Bruh.
It’s the direct analogy of a CEO who comprehends how an entire organization operates without knowing every minute detail.
The main reason they would need anything at all – so no one is bullshitting you!
I like that emphasis: anyone can do general managing work, but to sense something shady?
Know limits, inefficiencies, friction points? You have to know the puzzle and how it connects.
The architect has a vision, understands constraints and possibilities, and can navigate complexity without getting lost in it.
See? Building, navigation and understanding are one area of thinking.
Applied to the whole society it is… common sense! Even if the term is too vague, cause it’s both common sense, critical thinking, understanding cause and effect, biases, balance of probability…
This is more general thinking than just knowing stuff. Not to learn a thousand facts. Those are irrelevant now.
Everyone is so focused on teaching a million facts but not on understanding how to verify and sense them being sus.
One example would be like physics and history. For physics you do not need to know all the formulas, you just need to know what the speed of light is, its limits, and how general relativity operates, but you don’t need the formulas or how to calculate any kind of crap in the first place.
You would need that only when you actually commit to the science work.
For history – same.
They all want you to memorize a lot of facts like when battle X took place. Who needs that in modern society when rules of war are rewritten every year?
But you do need to know how to understand a time in the past and how things were. What sentiment there was, what was likely, what was not, cause and effect of events – how it could be predicted, or was completely random.
Being able to do a bullshit check of any random historical fact using broad knowledge, before you waste time on actual fact-checking.
For example someone tells you that 200 years ago one country invaded another, when in fact they were on opposite sides of the globe, and you should not believe that – because you understand what geography is and what the limitations and capacities of that time were.
Or you don’t need to know how many countries the British Empire invaded – just to know what imperialism is and how it operated.
Being able to do a quick sweep to ask the right questions from the actual fact storage bank.
The manager. Everyone would be a manager.
Like in SWE, we no longer write code – you no longer do anything with your hands. Quite literally!
Even operating your computer would first be fully automated with voice commands and a local agent – and then even that would be replaced with neural interfaces.
Be it just writing prompts, issuing commands to other people, be it managing other people or autonomous agents, or being integrated directly into the system, being on the same level as artificial beings and communicating with them through your direct thoughts, it’s still the same.
99% of people would be doing some kind of managerial work – regardless of their field, hard science, soft science, any financial or entertainment stuff – in the end it would always be some kind of management of autonomous or semi-autonomous agents, biological or artificial.
Your job – giving directions, thinking what to actually do and doing a bullshit check.
Like a little CEO of anything. Every job would be reduced to being a middle manager.
Because if you really start to dismantle any part of anything – most of it is very similar to being a middle manager.
Yet we hold the assumption that all of that is still unreliable, not trusted, but still.
Even super duper ASI – you would not want to trust it, being too capable – so back to sus-check everything!
Not to be so pessimistic – it’s not about "mistrust of everything".
Rather, when we are making mistakes, are we even doing a good job? When you are the planner!
Having an army of perfect executors will always bite you with being too perfect.
This is where knowing systems on a high level becomes important – knowing the limits of yourself and others. When they are lazy, make mistakes you don’t see, or are straight up lying.
I think it would be true on all time scales – both now and in 500 years.
While we would rely on automated workflows 99% of the time – 1% is where we would focus. When scaling up – this is where our most creative work would be.
Unknowns would be the new principle of information distrust. But also classic misinformation remains, so you always benefit from this skill.
Diplomacy
Fits nicely in with the other two – while diplomacy is a part of management, this is separate!
Interacting with other managers, handling agreements, forming and breaking relations, even emotional work – would be monetized diplomacy of some sort.
It does not stop at others only. Diplomacy with others, with yourself – the most important of your life, with whom you argue every day anyway!
Where do relations with others and self start? At ethics and values. Knowing what you and others want, where to even move.
While the architect skill is about thinking in creativity and execution, this is understanding of self and your own place.
Test of time
One way to prove this idea is to dismantle any part and see if any stays relevant.
Specific coding languages? Gone. Architect thinking stays.
Specific jobs? Gone. Management stays.
Specific social structures? Change. Diplomacy/EQ stays.
Society evolved to the point of full utopia, where you do nothing? Diplomacy with self.
You are now absolutely alone – same.
An entirely new threat emerges, fully unknown and unthinkable – the architect has to make an analogy and connect the dots.
When writing this I imagined an example – you hop into a spaceship all alone and fly into Andromeda.
Arrive and establish a colony of yourself, and you have to use all three levels of thinking:
Being an architect – how you build your colony, how you imagine it, where and what ideal place, making plans now to be viable in the next 100 years.
Being a manager – how you manage your fleet of robots who build it, how you delegate.
And diplomacy with yourself – while you’re all alone, or with entities you create, or even new species – it becomes multi-entity real quick!
Rebuilding a society from scratch would be the ultimate test of it. You need all three to make it awesome.
Yet meta-thinking about applying those skills not now but later – is an education in itself.
When you learn – think – would it be relevant in 10 years? Everyone should be a futurist right now, or be forever behind.
Test of future
Speculation of why it all could be wrong (you may wanna skip it).
What breaks it? This quickly collapses into ‘what being alive is’, and yet a good point!
Aren’t we supposed to teach ‘how to be alive and function’ rather than knowing how taxes work (lol they don’t even teach that!)
It relies on some assumptions about humans and intelligence, but more about – will we even stay the same?
Cause the entire point of this so far is to be able to adapt to everything there could ever be.
Where does Architect break?
I think this one will stand as long as we remain human – curiosity and desire to understand stuff.
Even faced with something unthinkable – our goal shifts to grokking it!
Some may give up but not all.
In a condition where Human Comprehension <= System complexity, we will not rest until we get it all!
Might be a futile quest, but so far all our history is about pushing forward and unless we are completely wiped I doubt any fancy ASI will stop us.
As long as we care – no black box lasts forever.
Or, if the level of complexity becomes so high it’s impossible – like a complete loss of causality, when we no longer understand how any of the world works – I think we just stop functioning at all until we move a level down, where the challenge awaits again.
However – you may find a case where going a level down is no longer possible – anything quantum-related may be so. When the capacity to understand is broken at the design level.
Architect thinking doesn’t scale down – it evaporates. You become a shaman performing rituals at the altar of outputs you cannot connect to inputs.
But! A problem you cannot solve – one you can ignore!
There might be the case when all our problems are either solved or impossible to solve and we proved it – like fundamental physics.
Then we either stop and forget that systems thinking is needed entirely – or use the same skill to invent new problems! (we always do)
When does Manager break?
One case would be when agents are able to communicate faster, at the speed of light – in such a way that humans are no longer a bottleneck.
When agents can coordinate faster than a human can think, the human manager is not just inefficient; they are a liability.
Or when we design a system that no longer involves decisions – we are removed entirely.
So we switch into goal-setting and general thinking what to do! Or possibly to owning stuff: as in the case of crypto – you never manage transactions yourself, only own. Or decide what to even do with all those coins.
Remember – this assumes we still want things AND want to choose ourselves or as a collective.
Even if all so perfect to the point we figured out all and everything and there is no more work or decisions to be done – who would be managing their own life?
Where does Diplomat break?
One idea is at post-scarcity – which never exists. One way or another, some form of currency, some ‘want’, some limited resource remains.
Even so, all covered, you can do anything! What you do – you start with yourself, asking what do I even want?
Or when we become a hive mind? A skill as learned before may be completely irrelevant, but the principle may still stand. Even a collective has to operate on some agreement – if it’s a true collective and not an overmind with drones.
The only way it becomes irrelevant – at the complete loss of both personal and collective identity – is if you’re neither a person, nor a hive mind. But it’s basically death. A Matrix-style removal / hack of the reward system – but you just effectively stop doing anything!
And we are here learning things to do something at all – so it no longer applies. Nothing ever applies in this case!
One case may be – when we experience learned helplessness at civilization scale.
When teaching or learning anything does not matter anymore.
Skills and the method of teaching are atrophied, replaced with cognitive substrate.
Similar to reward hack? Or closer to the dissolution of identity via an overmind that decides all? Either way – effective death.
Or – if there is no longer ‘you’ but a part of a larger thing – then the same skills apply to the larger thing itself.
An overmind would have the same problems.
What if you are in a perfect simulation? Does it matter? If we know – we will try to break it, or prove it’s impossible and move on. We always do. That’s the architect in us, again.
Then there is the complete ontological shift – what if the substrate of reality itself changes in a way that makes "systems" an obsolete category? We discover we are in a simulation and learn to exploit the code, glitching reality itself. Or quantum many-worlds becomes practically navigable – you are constantly splitting into branches. I would assume this would either break us entirely, extinction, or require mapping the old to the new – and we have to connect the dots somehow. Same skill, just with a new surface to learn.
Or a collective complexity so dense we can no longer function, or it becomes irrelevant. If you have to communicate to 10^256 NPCs at once to agree on anything… We either start solving it as a problem, or each retreats into their own. But if no own – then it’s death, or the death of the larger collective at once. Either way, the diplomat’s job is the same: find the agreement, or admit there is none.
And when transparency becomes so great that diplomacy between each individual no longer matters… All is deterministic enough where no disagreements can form – you retreat to self!
Where else would there be to go? The skills don’t break – they just have no one to play with. (Go invent yourself a friend)
It’s simple, actually.
And that’s the whole thing, really.
Architect, manager, diplomat – at the bottom it’s all decisions. What to build, what to delegate, what to negotiate, what to even want.
Yet – 3 words, infinite complexity behind.
And here’s the thing about that – we will keep handing these to machines, more and more, layer by layer. We’ll let them architect our schedules, manage our fleets, negotiate our terms. But every time we do, we just push the decision up one level – now you’re deciding which system to trust, what values it carries, when to overrule it…
The decision never leaves. It just changes shape. And I don’t think we want it to – not because we would be bad at giving it up, but because giving it up would not even be a thing we could want.
There is no one to do the wanting for us. Even you, alone on Andromeda, deciding what kind of colony to make – that is the whole skill. The machines can build it. The machines can run it.
But imagining, deciding, even as simple as picking a template – that would always be valuable.