Two years before ChatGPT, we told readers that learning to code was the safe bet; generative AI then hit writing and translation first, and soon reached routine coding too, so here is the question again with a more honest answer.
In October 2020 we ran a guest post called How to Know If Your Skills Are Applicable to the Future of Work. The question was a good one. It is an even better one in 2026.
The answer, though, needs redoing. The post was written about two years before ChatGPT arrived, and its safest bet, learning to code because "coders and programmers will be needed to direct these technological replacements", turned out to be close to one of the first things generative AI started to change. It also contained a plain factual error, and it leaned towards coding bootcamps, which makes more sense once you know its author co-founded a bootcamp marketplace. So here is the question again, with what we know now.
What we said in 2020
- A Federal Reserve Bank of St. Louis study found "nearly two robots per thousand human workers", and future skills would be "centered more around technology".
- If your work already deals with technology, your skills are probably safe. Programming was one of the most sought-after skills, and "coders and programmers will be needed to direct these technological replacements".
- Soft skills carry over between industries more easily than hard skills.
- Your industry may already have been disrupted. Accounting and finance, for example, had already been changed by spreadsheets and software.
First, a correction
The original said that "coding and programming only really emerged in the last decade". That is simply wrong. Programming dates back to the 1940s and 1950s. We should have caught it at the time.
What happened next
The "safe" knowledge work was hit first
The 2020 post assumed that robots and AI would replace other people's work, and that people working directly with technology would be the ones directing it. Generative AI, from late 2022, went the other way. It reached knowledge tasks first.
Research on freelance platforms found that demand for writing and translation work fell after ChatGPT was released (WashU Olin, ScienceDirect). These were computer-based, skilled jobs, not the physical work most people expected automation to hit first.
Coding did not stay the safe harbour
"Learn to code" is the most dated advice in the original. AI now automates a lot of routine coding as well as routine writing. That does not make technical skill worthless. It means that knowing how to write code is no longer, on its own, the guarantee we made it sound like.
The premium moved to working well with AI
The skill that stands out now is not a particular language. It is being able to use AI tools well. Upwork's 2025 Future Workforce Index reports that 54% of freelancers describe advanced AI proficiency, against 38% of full-time employees (Upwork). The people who work for themselves, in other words, seem to be adapting fastest.
The robots, for the record
The original's robot figure came from the St. Louis Fed. The International Federation of Robotics puts the global average at 132 robots per 10,000 manufacturing employees, with a record of 267 in Western Europe (IFR World Robotics 2025, released April 2026). That is a different measure, counting factories only, so it is not a like-for-like comparison. But it is a reminder that robots on factory floors were never the whole story. The bigger change came through software.
So, how do you know if your skills will last?
Here is how we would answer the question today. This is practical advice from us, not a forecast, and we are not selling any course.
1. Look at the tasks, not the job title
The 2020 post asked whether your skills were "specialized in tech". A better question is which of your day-to-day tasks a capable AI tool could already do a first draft of. Writing, translating, summarising and routine code were hit first. If much of your week looks like that, it is time to change how you work, not to panic.
2. Learn to use the tools on your own work
The survey above suggests that hands-on AI skill is where the gap is now. You don't need a bootcamp for that. Take one real task from your own week and practise doing it with an AI tool, then check the result properly. The checking is a skill too.
3. Keep your soft skills sharp
This is the part of the 2020 post that still holds. Judgement, communication, working with clients and colleagues, and knowing what a good result looks like still travel across industries. If anything, they matter more when a machine can produce the first draft.
4. Ask whether disruption has already happened
The original made a good point here: many industries were changed by technology long before anyone called it disruption. The same is true now. If AI is already in your workplace, the question is not whether it will arrive, but whether you are one of the people who know how to use it.
5. Don't bet everything on one skill
In 2020, coding looked like the safe bet. In 2026 it looks less certain. Whatever looks safe today may not in five years. A mix of skills you can move between jobs is a better hedge than one skill you hope will never change.
The honest scorecard
- Future skills will centre on technology: partly true. The direction was right.
- Coders will direct the robots, so learn to code: aged worst. AI reached writing early, and routine coding soon after.
- Coding only emerged in the last decade: wrong.
- Soft skills transfer across industries: still valid.
The question in the title is still the right one to ask yourself, maybe once a year. The answer just keeps moving.
Written 3 November 2026 from the research behind the archive's 2026 updates (checked 30 September, re-verified 3 November 2026). Sources are linked in the text.