James Needham.
“There’s a difference between knowing the path and walking the path." — Morpheus, The Matrix
We have spent much of the past two years asking what AI will do to our jobs.
I wonder whether the bigger question is what it will do to us.
AI adoption across OECD businesses has almost tripled, from around 7 per cent in 2021 to 20 per cent in 2025. AI has moved from corporate curiosity to an essential co-worker. To borrow from Pauline Hanson's tired political playbook, AI really is the foreign worker coming for your job. The difference is you cannot stop the boats when they arrive through the cloud.
The more confronting number comes from the World Economic Forum. By 2030, employers expect 39 per cent of the core skills we use at work today to change.
Forget predicting which jobs disappear. In five years, two fifths of the contents of our professional toolbox may need replacing.
That is quite a lot of life about to get rearranged.
At Untangld, AI is already foundational to how we work. We use it to devour information, interrogate markets, find themes, challenge hypotheses and travel through huge amounts of material at a speed that would have seemed faintly ridiculous five years ago.
Like many businesses, we are learning that access to information is becoming less of an advantage. Everyone can increasingly reach the same research, the same signals and often the same neatly packaged answer.
Which makes what happens after the answer much more interesting.
A machine can surface a pattern but having sound judgement begins with wondering why it exists in the first place.
Malcolm Gladwell gives a wonderful example in his book Outliers. Korean Air had accumulated a disastrous safety record through the 1990s. The crashes, cockpit recordings and flight data were all visible. However, Gladwell became interested in something less obvious to the numbers: the culture inside the cockpit.
Junior crew members could see danger developing but often softened their warnings to senior captains for fear of breaking the chain of command. As a result, the hierarchy shaped language and warnings became soft suggestions. The revealing insight sat beneath the data, buried in culture and human behaviour.
AI is extraordinarily good at hunting for patterns across mountains of information. What remains harder is knowing which pattern deserves our attention in the first place.
Gladwell stayed with the mess and kept digging past the convenient explanation until a better question emerged.
That appetite to remain uncertain long enough to find the question nobody else has asked may become one of our most valuable skills.
Then comes another very human act.
You have to make someone care.
Much of our value comes after the analysis from shaping insight into a narrative, reading a room, knowing when an argument has landed and when it has disappeared over everyone's heads. It comes from judgement, empathy, curiosity and occasionally having the confidence to say that the beautifully produced answer in front of you is wrong.
But there is another consequence of all this that troubles me more.
How do people learn to develop that judgement in the first place?
The uncomfortable truth is that much of the grunt work AI is removing was also the apprenticeship most people had to go through, especially if you were born last century.
The junior researcher who spent hours pulling apart the data eventually learned which number smelled wrong. The graduate building version fourteen of the deck slowly learned what made version fifteen persuasive. The young consultant sitting nervously in the client meeting was accumulating thousands of tiny observations about influence, politics and people.
None of those tasks looked particularly profound at the time. Yet together they built the pattern recognition, confidence and judgement we later came to call experience.
If we strip away too much of the work then we may accidentally strip away the pathway through which expertise is formed.
We should absolutely remove pointless labour. Nobody needs to romanticise hours spent formatting slides or manually compiling information a machine can process in seconds.
But efficiency cannot become the only objective.
If two fifths of our skills are going to change, teaching people the next set of skills will never be enough. We need to build people who are good at becoming good at new things.
That means every worker becoming fluent with AI through daily use. It means companies redesigning junior careers around exposure, experimentation, responsibility and mentorship. It means putting young people closer to the decisions and conversations from which judgement is formed.
And it means thinking differently about school.
The children entering classrooms today will grow up with extraordinary machines beside them. Machines that will remember more, calculate faster, write more fluently and retrieve more information than any human ever could.
We can't shy away from that and trying to compete on those terms feels futile.
Our job is to develop the things that help them decide what to do with all that intelligence.
We need to teach our kids about curiosity, critical thinking, creativity, communication and empathy. They need to build the confidence to respectfully challenge an answer. They need to develop the resilience to watch something they spent years becoming good at suddenly become cheap, then move again.
The tricky bit is we are conditioned for continuity. Our careers, qualifications and identities have traditionally travelled along reasonably predictable rails.
That world is fast fading.
The next generation may look back on the idea of a "career" as we look back on a job for life. They may move between problems, projects and industries, assembling new skills as quickly as old ones become cheap.
So the advantage now shifts to a new set of skills.
It will come from knowing what to ask and when to doubt the answer. How to learn something new quickly and how to spot what everyone else has missed. Then most importantly, how to bring other humans along with them.
The super power beneath all of these skills is adaptability.
In an age of abundant intelligence, the enduring advantage may be the ability to keep becoming someone new.
We just need to use our best judgment to discover what that 'new' is and then walk that path.
See you in the Matrix.
James Needham | Founder Untangld