Why Robots Will Take Your Jobs… and Why They Won’t


Learn about the rise of Artificial Intelligence and Machine Learning and the types of jobs they will take from you.

Shhh… do you hear that buzz? It’s Artificial Intelligence (AI).

Robots are putting fear in the lives of many Americans. (72% according to Pew Research) Now you might be thinking it’s about the robot apocalypse portrayed in most sci-fi films, but their top concern is that robots will take over their jobs. Modern-time robots have been taking over jobs in customer service, manufacturing, and transportation, but this isn’t a recent phenomenon. In fact, robots have been taking jobs away from humans for generations now. Think back to the 1950’s, when phone calls were made with switchboard operators physically inserting plugs into the appropriate jacks. Or when there used to be elevator lift operators to take you to your floor. These jobs have since been computerized, eliminating the need for human interaction.

Switchboard Operators Connecting Phone Calls (1952)

So if this has been happening for over 50 years, why all the buzz now? It’s because we are beginning to see the true potential of AI come to life. AI has evolved and most recently expanded to different areas of our personal lives. From Amazon Alexa to driverless cars to pizza-making robots, the possibilities are endless.

Check out my other blog about driverless cars: “Self-Driving Cars are on the Rise, But Can They Really be Driverless?

People may not even realize they’re using AI, but even the autofill function in your Google search and your shopping behavior on Amazon is using AI to provide predictions and suggestions. But AI can only be as smart as humans right? Wrong. Machine Learning (ML) gives the robot the ability to learn beyond the intelligence that was programmed into it. That means the robot can be better than the very programmer that created it. Nervous yet? Before we jump into the types of jobs robots will take, let’s get a brief understanding of the difference between AI and ML.

AI is Getting Deeper

Artificial Intelligence is the general concept of mimicking human abilities to perform narrow tasks in an intelligent way. These abilities are typically hardcoded to the machines (line by line) and limited to those task-specific algorithms.

Machine Learning takes AI one step further with the idea of having machines teach themselves. Rather than providing task-specific algorithms, machines are given access to data that allows the system to continuously train and improve itself. ML’s capabilities are limited to the specific goals that they were designed for. This means they are not able to learn new goals unless they are designed/coded to do so.

But wait, there’s more!

Deep Learning (DL) is a subset of ML with the goal of simulating human decision-making by leveraging artificial neural networks (ANN). In other words, DL enables machines to use layers of logical reasoning to draw conclusions, similar to how a human would. Let’s look at an example of how this works. Imagine you’re tasked to determine if a picture is of a cat. What do you look for? You may look for pointy ears, whiskers, long tail, etc. How can you tell it apart from cheetah or lion? You’d probably check for spots, facial structure, or body shape. DL follows the same feature-set approach, working through layers of logical steps to draw a conclusion.

“This is a Cat!” — Deep Learning Robot

Here’s a great blog post by Analytics Vidhya to learn more about Deep Learning vs. Machine Learning.

Types of Jobs Robots Will Take

Economists have divided types of jobs into four basic categories:

  • Routine Manual: Predictable and repeatable physical tasks
  • Routine Cognitive: Predictable and repeatable cognitive tasks
  • Non-Routine Manual: Non-predictable physical responsibilities
  • Non-Routine Cognitive: Non-predictable cognitive responsibilities

Sooner than later, robots will take over all jobs defined as Routine Manual and Routine Cognitive. Think of tasks performed by telemarketers, receptionists, office admins, customer service reps, analysts, factory workers, line cooks, cashiers… These are jobs that repeat the same mundane tasks over and over again and can be easily replaced with robots. Add in the fact that robots will be cheaper (over time), be more productive, and perform more consistently, and you have yourself the perfect human replacement.

As ML and DL continue to improve, you will see Non-Routine jobs be taken over as well. Some optimists believe that by 2060, robots will be able to do any tasks that a human can. Even jobs that require the most cognitive and creative skills, such as art, management, legal, engineering, education, and music will be at risk. Can you imagine a robot making the next #1 hit song? Well, that’s already in the works: Robot Composes and Performs It’s Own Music

These four types of jobs can be chronologically ordered to depict when robots will supersede humans. The image below shows a high-level representation for the increase in types of jobs taken by robots over time.

Increase in Types of Jobs Taken by Robots Over Time

What Does Ricky Think?

At some level, every job can be done by a robot. The question is not if robots CAN take the job, the question is if they WILL take the job. Jobs that require building complex relationships or a human touch will likely not be taken by robots. These may include mental health physicians, personal trainers, nurses, teachers, motivational speakers, therapists, masseuses, business managers, etc. Most people will still enjoy interactions with actual humans, which will safeguard these types of opportunities.

Another area likely to stay with humans are professional athletes. At this point, its safe to say there will be robots that can shoot farther than Steph Curry or throw better than Peyton Manning (I’m a Colts fan). But the passion for human competition and athletic excellence will drive the demand for humans to continue being professional athletes. Although I am excited to see robot boxing matches in the future. Real Steel?

Boxing Robot (Image from Real Steel, 2011)

Lastly, politicians are likely to be safe from robots taking their jobs. Even when robots are fully capable of making risk-adverted decisions just as well (or even better) than today’s politicians, they will not replace them. Most politicians will fight against that and even have the power to create legislations to protect the very seat they sit in.

Now if you’ve read up to this point, you might be saying: “Oh no! Only a small percentage of jobs will be left for humans.” While it may seem that way, the more appropriate thing to say is “Only a small percentage of TODAY’S jobs will be left for humans.” Robots will continue to evolve and improve — but so will humans. Could we have predicted 50 years ago that we would have millions of engineers building mobile apps today? Probably not. AI technology is intended to complement humans and give them the opportunity to perform higher-level tasks. These opportunities will lead to more imaginative ways to create new jobs and advance our way of life.

My key takeaway for you is: Stay hungry, challenge yourself, and learn something new everyday. Your skills today may be commoditized tomorrow.


Did you find this useful? Please recommend or share, and feel free to hit the clap icon. 👏🏻

Source: Deep Learning on Medium