There are decades in which nothing happens, and then there are weeks in which decades happen… (Lenin)
Much has been said about the future of skills, but the focus is often on a distant future, slowly coming about. That future is here. And the skillsthat got us where we are today, won’t get us where we need to go tomorrow.
“Skills” refer to an ability to perform a task in a knowledge-intensive industry.
TLDR
- The rate of change in knowledge-intensive industries is outpacing the education system's ability to deliver the necessary skills.
- The average time a skill stays relevant is now around four years, down from over ten years in the past, and this time is continuing to shrink.
- Qualifications are losing their value, and the model of "learn first, apply second" is becoming obsolete.
- The ability to adapt and learn as you go along will be increasingly valuable.
- Artificial intelligence is changing the skills that provide value in many industries by replacing technical skills with an ability to guide AI to produce quality results.
- Critical evaluation skills will be important in working with AI to produce effective results.
- The world of software engineering provides a window into the future of what other knowledge-based skills will look like, with a focus on constantly updating skills and manipulating AI tools to do specific tasks.
- Key skills that are important for evaluating software engineering candidates include the ability to collaborate and the ability to learn.
Qualifications are losing their value
The rate of change is outpacing our education system’s ability to deliver the skills we need. The average time a skill stays relevant used to be over ten years. Today (2022) it is four, down from five in 2017. Soon, it will be shorter than the time it takes to get a degree.
We keep applying plasters to the problem, decreasing the time required for qualifications and inventing micro-credentials to recognise ever smaller units of skills. But they are temporary solutions. The value of a qualification is only as good as the time for which it provides you withan advantage in the workforce. As that time keeps shrinking, the value of qualifications shrinks with them.

There are decades in which nothing happens, and then there are weeks in which decades happen… (Lenin)
Much has been said about the future of skills, but the focus is often on a distant future, slowly coming about. That future is here. And the skills that got us where we are today, won’t get us where we need to go tomorrow.
“Skills” refer to an ability to perform a task in a knowledge-intensive industry.
Qualifications are losing their value
The rate of change is outpacing our education system’s ability to deliver the skills we need. The average time a skill stays relevant used to be over ten years. Today (2022) it is four, down from five in 2017. Soon, it will be shorter than the time it takes to get a degree.
We keep applying plasters to the problem, decreasing the time required for qualifications and inventing micro-credentials to recognise ever smallerunits of skills. But they are temporary solutions. The value of a qualification is only as good as the time for which it provides you withan advantage in the workforce. As that time keeps shrinking, the value of qualifications shrinks with them.

The logical endpoint is that qualifications, as we know them, will lose their value. With it, the entire model of “learn first, apply second” will become obsolete. And as the value of any particular skill decreases, the value of your ability to adapt and learn as you go along will soar.
Artificial Intelligence is changing the game
Artificial Intelligence (AI) brings this ability to adapt into sharp focus. On the one hand, AI accelerates the pace of change even more - by changing how we work (and the skills required) in many industries and for many jobs at once. On the other, it changes the skills that provide value in the first place. Instead of doing the “hard” work of building a presentation, writing a new policy, or producing a new budget, the work will be about guiding artificial intelligence to do the work instead.
The technical skills required to do our jobs are rapidly becoming obsolete, replaced by an ability to guide AI in a way that produces quality results - and the critical evaluation required to iterate over those.
A request to “write me an IT policy” won’t get you anywhere near the result of a request to “write me a short IT policy, in the form of bullet points, for a small company that does most of their work from home. Make it strict, but allow people to install apps on the devices the company provides them.” - let alone what happens when you ask the AI to tweak the result and allow people to share devices within certain contexts.
And even though “give me a few examples of a sales pitch
for X” can provide interesting results, it becomes much more interesting when you realise you can tweak your question to ask: “write me a sales pitch for product X, highlighting value propositions Y and Z, personalised to brand A and contextualised to event B that happened
last week. Make it short, to the point and use a thoughtful tone of
voice”.
So where does that leave us with the future of skills?
Learning from software engineering
As is often the case, “the future is already here, just unevenly distributed” (William Gibson). I believe that the world of software engineering already provides us with a window into the future of what most other knowledge-based skills will look like.
Software engineering has been changing faster than
almost any other industry and constantly keeping one’s skills up-to-date has been part of being an engineer for a long time. The amount of libraries and tools software engineers have to know and be aware of every day is staggering. Below is a graph of the number of “npm modules”(libraries that provide certain capabilities which software engineers can use to build apps):

Source: http://www.modulecounts.com/
More than TWO MILLION modules exist, and that is just one development environment (node.js). We can draw two broad predictions from this:
We’re moving to a world where the time a skill stays relevant is getting ever shorter
A series of AI tools are springing up to do specific tasks for us, while our job becomes about manipulating and orchestrating them
To test these predictions, I asked a group of CTOs to rank the most
important skills they look for when evaluating software engineering
candidates and they came back with the following:
Ability to collaborate
Ability to learn
Ability to problem solve
The ability to code didn’t even make the top three!
So much of engineering today is about leveraging existing
technology, understanding how to use it, applying it to your particular situation and being creative in doing so. The ability to write code plays second fiddle to all of the above. The same will likely be true of many more jobs in the future.
Closing thoughts
In a world where the time any particular skill stays relevant is counted in months, not years and various AIs jump in to perform the “hard” tasks of preparing a budget, writing contracts, brainstorming strategies and writing code, the ability to learn, collaborate and problem-solve become the only things that matter. A junior finance professional with a great ability to learn, an open attitude to collaboration and great problem-solving skills will likely outperform the CFO with 20 years of experience and a degree from a well-known university that doesn’t (know how to) use the latest tools.
Note: As these new skills take top priority, it poses the question of how companies will measure them and the proxies they will use to hire great candidates. But that’s a topic for another day.
