Work that is difficult to automate

When we're talking about robots taking people's jobs, we're speaking of automation.

Mechanical automation, like car assembly lines, has been around for a while.

Software automation, also known as process or work automation, involves using code to automate tasks that humans would otherwise have to do, like creating an invoice in an accounting program.

Low-skill jobs, where 70% of the responsibilities are predictable physical and cognitive tasks, are straightforward to automate, especially as automation technology becomes cheaper than paying a human to do the same job.

• Outside an office environment: The jobs that will be affected are things like retail employees and warehouse workers.
• Inside an office: Jobs vulnerable to automation are data entry, filing, and document review.

Complex tasks that require creativity and other forms of higher-order thinking are very difficult to automate. The reason is that you need cognitive technology like AI (artificial intelligence) and automation together. At this point, AI is still limited.

Artificial Intelligence lacks social intelligence and human warmth. 

• There are concerns if AI can make ethically appropriate decisions.
• Concerns about bias in AI can make companies wary to use it for complex office tasks, like hiring. 
• There's also a shortage of automation and AI talent.

Depending on the pace of automation development, we could see some higher-order tasks being automated.

If you think about it, your job has probably already changed. You're not doing everything by hand. In terms of scale, the changes society experience now are similar to about 100 years ago, at the start of industrialization.

Many jobs now focus more on creativity, decision-making and other forms of higher-order thinking.

• By 2022, 62% (compared to 46% today) of an average business's data processing, information search, and information transmission tasks will be performed by machines.
• Machines will do tasks like communication, management, and decision-making to some extent.
• Entirely new jobs will likely emerge as technology isn't self-sustaining.
• Though technologies like automation and AI will displace 75 million jobs globally by 2022, they'll also create 133 million new ones. 
• A McKinsey analysis estimates that 8-9% of 2030's labor supply will be in roles that don't currently exist.

In advanced economies, up to a third of the workforce might need to learn new skills and find new occupations by 2030.

To avoid being out of work later, it is best to start upskilling now:

• Get better at working with data
• Learn a programming language
• Keep up with the best tech practices for your field.