Every wave of automation has quietly raised the value of something it could not do. Spreadsheets did not make accountants redundant; they made the accountant who could explain the numbers to a board more valuable than the one who could only produce them. The same pattern is playing out now, and it points to a set of skills that are easy to name and hard to build.
Why the soft skills argument has changed
For years, creativity, empathy and critical thinking were described as nice-to-have qualities that rounded out a technical profile. That framing no longer survives contact with the evidence. When drafting, summarising and formatting become close to free, the scarce inputs are the ones that decide what to draft, whether the summary is trustworthy, and what the organisation should do about it.
These are not personality traits. They are practised capabilities, and they improve the same way technical skills do: through deliberate repetition with feedback.
Three capabilities worth building
1. Framing the problem
Most poor outcomes trace back to a well-executed answer to the wrong question. Framing means slowing down long enough to state what decision is actually being made, who it affects, and what would count as a good result. It is unglamorous and it saves months.
2. Judging quality quickly
When output is abundant, the bottleneck moves to review. Being able to look at a draft, a model or a plan and locate the weak joint in under a minute is a real and teachable skill. It comes from having seen enough good and bad examples to know where things usually break.
3. Explaining across a gap
Technical work only creates value when someone who does not share your expertise can act on it. That means writing plainly, choosing the one number that matters, and being honest about uncertainty rather than hiding it behind confident language.
Making it part of the working week
Skills like these are rarely built in a workshop alone. They are built in the ordinary work, if the work is set up to teach.
- Write the problem statement before the solution, and ask a colleague whether it matches what they thought you were solving.
- Review one piece of someone else’s work each week and give specific, kind feedback. Reviewing teaches faster than producing.
- Rewrite one technical update for a non-technical reader. If it gets shorter, you understood it better.
Organisations can help by making these behaviours visible. If reviewing, mentoring and clear writing never appear in how performance is discussed, people will correctly conclude that only output counts.
Key takeaway
Human skills are not the soft edge of professional life. They are the part of the work that decides whether everything else was worth doing, and they respond to practice just like any technical skill.