It’s 2038. What kinds of jobs are available?
What will the job market for humans beings look like in 20 years time?
No one can predict the future, but I think it is possible to identify the kinds of jobs that will be important in an age of machines, and become more important the further we move into an age of machine intelligence.
My argument has three steps. First, proceeding by a process of elimination, I identify the kinds of human abilities that will still be required in the job market in 20 years time. Second, I drill down to reflect on a style of creative work that has become increasingly important in recent years and that (I believe) will continue to establish its importance in coming decades. Finally, I bring these lines of argument together to reflect, in broad strokes, on the kinds of jobs that will most certainly exist in 20 years time.
I will not try to scope out all the jobs that will exist in 20 years time (this would be impossible). I simply intend to pinpoint a set of jobs that will be necessary to enable business organizations to thrive and survive in the hyper-kinetic, data-driven, machine learning environments of the future.
Future-focused leaders should be trying to build these jobs into their organizations or to consolidate them if they’re already there. Future-focused job-seekers should be trying to skill-up in these forms of work. The good news is that as human beings, we are innately well-suited to do this.
What machines can and can’t do
There is an obvious way to figure out what jobs won’t be around in 20 years time. Identify all the jobs that robots and machine learning systems are already doing and subtract them from the jobs of the future.
These are jobs based in:
- Algorithmic calculation
- Formulaic, step-by-step processes
Jobs being taken by machines today include:
- Data research and analysis work
- Assembly-line and factory work
- Phone operators, telemarketers and receptionists
- Truck and taxi drivers
- Accountants
- Bank tellers and clerks
The list goes on. New jobs are being added (and taken away) every day.
Identifying jobs that machines can do today indicates the kinds of jobs that won’t be around in 20 years time. But it doesn’t tell us what kinds of jobs will be around in 20 years because most of them haven’t been invented yet.
To determine what kinds of jobs will be around in 20 years time, we need to think about the kinds of jobs that machines can’t do, or can’t do well, at least. We need to zero in on jobs that require uniquely human abilities — abilities that can’t be replicated by machines at current levels of machine intelligence and in the foreseeable future.
Here’s six human abilities that currently can’t be replicated by machines:
The ability to empathize with human beings
Empathy means ‘feeling with’ — feeling the same thing as another human being and expressing this in an intimate way. A machine can learn to fake this ability, but this would be a violation of trust (see below).
The ability to make a person feel acknowledged
Ever tried standing alone in a room, delivering a convincing monologue to a camera? It’s incredibly hard to get into the spirit of the monologue because there’s no person there responding to your words. Somehow, being acknowledged by another human being makes us come alive. It brings us out of ourselves and enables us to perform to our potential. It doesn’t matter how human-like a machine looks or behaves, if we know it is not empathizing with us, we might as well be alone.
The ability to make a person feel cared for
This follows from the two principles above. Machines can be designed to look after people, like support bots designed to ‘care’ for the elderly. But it’s impossible to design a machine that makes a person feel cared for, assuming that the person requires a genuine sense of empathy and acknowledgement. It is possible for a machine to fake empathy, acknowledgement and care, but this again risks violating the condition of trust. An example is Joi, the virtual girlfriend in Blade Runner 2049. K is fond of Joi, who calls him Joe, but his illusions are shattered towards the end of the film when runs into a promotional version of Joi who uses similar lines and also calls him Joe.
The ability to think critically about human life and society
If the abilities above are a litmus test for baseline human nature, the ability to think critically about life and society is the threshold step into human adulthood. I doubt that machines will ever do this well. The reason why has nothing to do with the complexity of the task (machines are already more capable than humans to gather and process all the data required to think critically about life and society), but because machines lack the empathetic, intuitive hardware required to process this data in order to produce emotionally and intuitively-resonant insights relevant to human beings.
Machines can be intelligent but they cannot be wise. A machine can be a surgeon but it will never become a philosopher.
The ability to establish trust
Machine learning is reliable, which is why we trust machines to perform algorithmic calculations and deliver on tasks every day. Imagine, though, if you walked into work tomorrow and found a robot employee sitting at the desk beside you. You might learn to trust the machine to do its job, but would you trust the machine itself? Would you trust something that was incapable of empathizing with you, that couldn’t acknowledge you as a person and make you feel cared for, and that was incapable of thinking critically about your life and society?
You would? I wouldn’t.
The ability to create art
This is the biggie. Machines can create stunning images through pattern recognition and algorithmic design. But a machine can’t create art because art involves a set of uniquely human responses to life and the world. This is what makes art, art — the fact that we can look at a work and feel with the artist, appreciating that they’ve seen, heard, and cared for other human beings, and reflected critically on human life and society.
This unique capability — our ability to create art — is a great place to start reflecting on what the job market is going to look like in 2038. The most important jobs for human beings in an age of machine learning will be jobs that have an artistic component, particularly jobs that involve applying empathy, critical thought, and artistic creation to decision-making.
Machines can crunch data, but we’re a long way from creating a generation of machines capable of critical questioning and rapid sense-making in the service of generating meaningful novelty—in a word, innovation.
Learning loops and empathetic connection
To round on my central insight, I need to say a few words about new forms of job design in the tech industry and companies that are engaging with it.
It is often said that software is eating the world, but it’s less often noted how the practices and mindsets of software development are eating the workplace. This is what is happening as large companies come into contact with the tech startup industry and reorganize themselves to operate with the agility and speed of the small companies they encounter there. Thousands of companies today are skilling up on the innovation practices associated with the tech industry, like agile management, lean startup method and design thinking.
These companies are creating a new generation of jobs that are quite unlike the jobs that existed 10 or 20 years ago. These jobs are highly collaborative, intrinsically creative, and demanding. Whereas traditional jobs required employees to execute of a regimen of prescribed tasks, these new jobs require employees to autonomous identify problems and generate solutions in collaboration with coworkers and, not infrequently, customers too.
All of these jobs involve an element of hacking. While there are plenty of criminal hackers, hacking itself is not a crime — it is a problem solving activity. Hackers solve problems by building things, testing them, learning from the results and trying again. This is how black hat hackers break through security systems. It is also how designers hack product ideas, entrepreneurs hack new business models, and developers build products, functional piece by piece.
The problem-solving process looks like this:
The art of innovation
So what will the job market look like in 20 years time? We know that machines will do most of the menial, formulaic, process-driven work. It’s pretty clear from advances in machine learning that machines will do most of the straightforward decision-making work undertaken by managers too.
What machines won’t have mastered is the art of rapid-fire innovation — the ability to see a problem or an opportunity, hatch a hypothesis for a solution, prototype it, test it, learn from the results and improve the model at pace. Machines can’t do this because they don’t have human powers of intuition, empathy, and engaged, critical reflection. They can’t create art. This is the major difference between human beings and machines.
The most important jobs for human beings in the 21st century will leverage our artistic abilities. These jobs will involve working with humans and machines to create amazing innovations. These jobs already exist today (in prototype form) in the tech and creative industries. Anyone who builds a prototype, tests it, and reflects on what s/he has learned engages in an artistic activity that will be vital in the workplace of the future. S/he leverages uniquely human capacities, critically reflecting on human life, and empathizing with other people, to create something that didn’t exist before.
Leaders should be building these kinds of artistic capacities into their companies right now — not just to help their employees engage and master new forms of work, but to prepare their organizations to engage the future, where machine learning and AI will drive businesses at breakneck speed, ramping up the scale and stakes of competition to unbelievable heights.
In another post, I will consider what kinds of companies stand to exist in 20 years time. These will be agile organizations quite unlike the arthritic bureaucracies that dominate the business world today. The heightened levels of competition, uncertainty and change in an age of machine learning demand that every product, service, and business process will need to become intelligent (meaning aware of and responsive to context) and agile (meaning adaptive to circumstance). Machines will play a role in this, obviously. But human beings are also required to make important decisions about things like customer preferences, and the impact of change on human life and society, as well as to provide the kind of creative input that machines simply cannot offer.
We sometimes think of innovation as a technical process. But innovation is made of human beings. So long as art, empathy, critical thought and creativity are required for successful innovation, there will always be jobs for humans.
Sharing is caring. If you like this post, please pass it on.
Tim Rayner is the author of Hacker Culture and the New Rules of Innovation (Routledge 2018). He teaches at UTS Business School in Sydney.
