Which human skills still count when AI does the analysis?
You have probably noticed it already. The analysis you used to set a week aside for is now done in an afternoon. The report drafts itself. The first version of the plan is ready before your coffee has gone cold.
And then you are sitting with the client. The numbers hold up, the recommendation is clear, everyone nods. Two weeks later you hear they went for the other option.
All that speed has not made that last part a day shorter. If anything, it weighs more. When everyone has the same analysis within reach, the difference lies in what happens next. We usually call that part soft skills, or human skills.
Human skills, often called soft skills, are the skills that make your work land with other people: explaining, earning trust, having a difficult conversation, getting a team or client on board. Now that AI does more of the analysis and execution, they decide whether anything happens with your work. They have one thing in common: you need to know how people choose. That is influence.
What are human skills, and why is work moving that way?
Human skills are the skills you need the moment another person is involved. Communicating, working together, persuading, listening, dealing with resistance. They are called soft because they are hard to measure. For anyone who has already mastered their field, they are the hard part of the job.
Work has been moving in this direction since long before language models. Harvard economist David Deming studied how the American labour market changed between 1980 and 2012.[1] His finding: the share of jobs that require a lot of social interaction grew by nearly 12 percentage points in that period. Jobs heavy on maths and light on contact shrank by 3.3 percentage points. Many technical occupations were among them.
Look at those years again. This happened when a computer could not yet write a paragraph.
Deming's explanation is a sober one. People who coordinate well with others make working together cheaper. The work is divided better and less gets lost along the way. I expect AI to carry that movement further. The more calculation a machine does, the larger the part of your day that consists of people.
Which soft skills remain when AI does the analysis?
Four. And you know all four from your own diary.
The first is explaining what the machine has worked out. That sounds simpler than it is. Berkeley Dietvorst, Joseph Simmons and Cade Massey of the Wharton School had people choose between a human's forecast and an algorithm's.[2] Those who had seen the algorithm make a mistake often dropped it afterwards. Even when it forecast better than a human. They called this algorithm aversion. So the output of a model is not trusted for free. Someone has to stand behind it and be able to say what it rests on and where it could go wrong.
The second is earning trust. In 1995 Roger Mayer, James Davis and David Schoorman described what trust rests on.[3] In their model a person looks at three things: whether you are able, whether you mean well by them, and whether you stick to principles they find acceptable. Only the first is about expertise. The other two your client reads from how you behave, and there is nowhere they can look those up.
The third is the difficult conversation. The budget that is not coming. The colleague who has been underperforming for a year. The advice the client was not hoping for. A model can help you prepare that conversation. It just is not in the room when the other person goes quiet.
The fourth is getting people on board. A team that has only just got used to the last change. A board with three other priorities. A client who says yes and does something else.
Machines can take over all sorts of things, but not the human side of getting people on board. Influence, in other words.
What do these four soft skills have in common?
On a list they sit one below the other, as four separate topics. Look a little longer and you see they come down to the same thing.
In all four cases someone else has to choose something. Your client chooses whether to believe the outcome. Your sponsor chooses whether to trust you. Your team member chooses what to do with your message. Your team chooses whether to come along.
And people choose differently from what you would expect if you only look at the content. They hold on to what they have. They watch what others do. They want to keep the feeling that the decision is theirs. A good argument often does not get through that, and you can read why in why facts and arguments rarely persuade.
If you know how people choose, you can take it into account. That is influence.
That is why influence is the missing layer for today's professional. You have the expertise. You have the analysis, faster than ever. And between that analysis and the result sits a person who decides.
The word sometimes brings tricks to mind. I mean something else. Influence is understanding what someone wants to achieve and what is holding them back. Then you make it easier for them to make the choice that helps them forward.
How do you learn to influence?
Influence can be learnt. There is a science underneath it and a method on top.
The science is behavioural science: decades of research into how people decide. The method for applying it is called Behavioural Design. At its heart is the SUE | Influence Framework©, which builds on the force field analysis of psychologist Kurt Lewin.[4]
The model starts with the person and with what they want to achieve. Then it maps four forces. Pains are what bothers someone now. Gains are what the new situation gives them. Those two push towards change. Comforts are everything that is pleasant about how things are now. Anxieties are the doubts and worries about the new. Those two keep someone where they are.
The core: change happens when you strengthen the first two and weaken the last two. Most plans, recommendations and presentations only do the first.
Take the client from the opening. Your report was full of their problem and of what your solution would deliver. What kept them where they were did not appear in it. I have worked out how to handle that in an advisory conversation in why clients ignore advice that is right.
The same four forces are at play with your team, with your board and with the colleague in that difficult conversation. So you learn one way of looking, and you use it wherever someone else decides.
What does this ask of senior professionals and managers?
The further you get in your field, the less your work consists of the field itself. A senior professional is judged on what the client does with their advice. A manager on what the team does when they are not there. Both get their results through the choices of others.
This group has usually done the courses already. Presenting, negotiating, giving feedback. What is still missing is the insight underneath: why it works with one client and not with the next.
You can read that insight in an afternoon. It only becomes a skill when you apply it to your own client, hear what happened, adjust and do it again. Why that so often stalls after a course is covered in soft skills training that sticks. Our philosophy is not to sell training, but to really develop skills.
The organisations that make the difference have professionals who can do this.
Frequently asked questions about human skills and AI
What are human skills?
Human skills, also called soft skills, are the skills you need in contact with other people, such as explaining, listening, persuading, working together and dealing with resistance. They sit opposite hard skills: the subject knowledge and techniques of your profession. For experienced professionals, human skills decide whether anything happens with that knowledge.
Which soft skills matter more because of AI?
Four: explaining what a model has worked out, earning a client's trust, having a difficult conversation and getting a team or client on board. These are the moments when another person makes a choice. That is where the work that remains begins once analysis and execution speed up.
Can AI take over soft skills?
AI can help you prepare a conversation, rehearse an explanation or think up objections. The choice itself is made by a person, in a conversation with another person. That part of work has been growing for a long time: economist David Deming showed that the share of jobs with a lot of social interaction in the United States rose by nearly 12 percentage points between 1980 and 2012.
What is the difference between influence and manipulation?
Influence is understanding what someone wants to achieve and making it easier for them to make the choice that helps them forward. Manipulation is getting someone to make a choice that harms them, or one they would never have made with full knowledge. The test is whether you can explain openly what you are doing.
How do you develop human skills as an experienced professional?
Start with what they have in common: knowing how people choose. The SUE | Influence Framework© maps which forces move someone and which hold them back. After that you practise on your own client, team or board, because a skill comes from applying and adjusting.
The part that remains
Over the coming years the doing will get faster, cheaper and within everyone's reach. It buys you time. And it takes away the excuse that the analysis was not finished yet.
A machine works out what the best choice is. Getting someone to make it is your job.