Over the coming years and decades, we’ll inevitably see a tremendous amount of change in the job market. Much of this change will be due to the rapid adoption of advanced automation technologies, like AI, by companies.
In these changing times, it’s useful for practically anyone to understand which skills will withstand the test of time during the age of AI. Not only will it help you develop a sense of what types of careers are more threatened by automation than others, but also identifying what these skills are will help you know which skills to double down on in your quest to becoming futureproof.
Let’s dig into three skills that AI won’t replace anytime soon.
Communication skills will always be invaluable. Sure, machines can communicate at a basic level today. If you ask Siri what the temperature is, it will communicate the weather in your location. But on a deeper level, communication isn’t just the exchange of facts but of more complex forms of information, such as ideas, concepts, imaginations, behaviors, and more.
So much of communication isn’t in what you say but how you say it. And sometimes, the meaning behind the words matters more than the words themselves.
Because machines aren’t conscious beings that relate emotionally to their experiences, robots cannot share powerful stories from the heart.
On a more technical level: one can argue that generative NLP models like GPT-3 are getting close to human-level communication skills. I’d say that the examples you see online are misleading – these models have no real comprehension of what they’re outputting. They’re essentially guessing what the correct text output would be, given the massive amounts of text that they’ve been trained on. It has a particular, narrow view of words and how they relate to one another. It lacks the experiences, raw emotion, and empathy that make up compelling storytelling. Sure, it could be an excellent tool for generating text that handles the more mundane part of the writing process or giving writers ideas for getting out of a writer’s block, but we won’t see an AI Hemingway or Jon Favreau anytime soon.
Summers-Stay of MIT Tech Review made an apt analogy to describe how GPT works:
"GPT is odd because it doesn’t 'care' about getting the right answer to a question you put to it. It’s more like an improv actor who is totally dedicated to their craft, never breaks character, and has never left home but only read about the world in books. Like such an actor, when it doesn’t know something, it will just fake it. You wouldn’t trust an improv actor playing a doctor to give you medical advice."
💙 Emotional competence
The most basic level of emotional competence is the ability to read emotions. As it stands, state-of-the-art sentiment analysis algorithms aren’t great at reading emotions through facial expressions or tone of voice. When it comes to more sophisticated emotion-related skills, such as intervening in emotionally complex situations or invoking particular emotions in others, AI isn’t close to replacing humans.
As automation technologies increase, it’s becoming increasingly clear that there is a need for ethics guidelines in the space. AI does not and cannot optimize for ethical decision-making unless we program it to do so. Qualified ethicists need to inject their moral values into machines responsible for making decisions that teeter on moral boundaries. There’s no clear definition of what is ethical or not, and even generally agreed-upon definitions can vary across cultural and geographical frontiers. However, we’ll need human expertise to help us develop standards for ethics guidelines that we can program into AI models.
Now that you know three skills that will be difficult for AI to replace, hopefully, you’ll feel motivated to double down on them. I should also mention that this is far from an exhaustive list. If you’d like to see more examples, send me a tweet about doing a Part 2.