AI is coming for our jobs – but which ones?
Is the job of a plumber or a carpenter more secure than that of an architect or a fund manager?
Charles Darwin University's AI expert, Associate Professor Niusha Shafiabady, takes a look at which businesses and industries will be affected most by AI.
Most of us, regardless if we are heading major corporations or living our quiet, everyday lives, want to know which jobs will be lost to artificial intelligence.
This is an important question, because it helps us make informed decisions about our careers.
AI’s strength lies in learning the repetitive patterns. Meaning, that if it wants to imitate someone playing ping pong, it should have seen many samples of how people play it. For the system to learn the patterns, we show it roughly 70% of the recorded ping pong games, which are our samples. This is called training the system. After the AI has been trained, we want to see how well it can imitate playing tennis.
Now, we show the 30% of the samples that it hasn’t seen before and check its performance. This is called testing.
Here, our aim is creating the AI system that best imitates a ping pong game – without knowing anything new about it. If we have good data, meaning that the samples we used to train our AI are professional games, and we train our AI well, it can imitate ping pong games very well. We call this system one with a high accuracy.
To answer the question on which industries will be affected by AI first, we should look at the tasks that can be done programmatically, because software is quite cheap to produce. (Not to maintain! But that is another story) We should also look for tasks that are based on repetitive patterns – like a secretary, whose main job is answering the phone. If, in an industry, we have those types of tasks, the jobs associated with them will be displaced relatively quickly.
Let’s look at the financial sector. A fund manager’s job will be displaced if we know their trading rules, because we can program it to a computer and a computer program can imitate their tasks.
Now, let’s look at the construction industry and explore the tasks related to a carpenter and an architect. Many architects have been using computer software programs in their day-to-day work for some time. AI developers can use their logic and design patterns to create smart algorithms to potentially imitate the architects’ tasks.
Eventually, this will lead to displacement of jobs for architects in the future. Comparing this to a plumber or a carpenter who is working in the same industry, we see these can’t be done by computer software.
We can have robots do a plumber or a carpenter’s job, but, as mentioned before, producing software is cheap. Since making robots is not as cheap as producing software, a plumber or a carpenter’s job will be less likely to be displaced in the near future by technological advancement brought about by artificial intelligence.
If we apply the same principle to any industry, we can foresee what areas will change and what jobs will be affected faster than the others.
The next question is more difficult: What are the implications of these changes for the society, economy, and businesses?
If we allow the technological tools to take over, let’s say we will end up with 50% job displacement. Let’s be positive and imagine that the governments have the resources to pay for job losses – which the pandemic showed they can’t do it without serious consequences to the economy.
If we allow the technology to take over people’s jobs to a great extent and half the population become unemployed, many things in the society, business environment, and communities will change, and the economy will be affected.
This is the decision we need to make today as a nation: Are we ready to move to the next level with these technological advancements?
If we aren’t ready, how can we prepare ourselves for what is coming?
Have your say