What is the Automation bias?
The Automation bias is a cognitive bias that occurs when people imagine they can create an automated system to replace a human.
Why is this bias dangerous?
This kind of automation bias can jeopardize jobs and put lives at risk by introducing more problems than it solves. For example, the automation will lead to increase risks as users won’t be as prepared as they think for their next interaction with the product.
Why do people use this type of bias?
People tend to have an optimistic view on technology that often goes far beyond reality. In addition, it would be easier for businesses to run tasks automatically instead of hiring employees so it could save money and benefit shareholders in the long term too. Other reasons include: a lack of awareness, a lack of education and inability to make sense of complex systems.
How do people use this bias?
Many times have we seen how new technologies have been designed with the user in mind first. That is, instead of putting themselves in the shoes of the person who will be using it. The best example for automation bias was a NASA research that wanted to create a machine that could detect early signs of cancer cells after surgery where small samples were taken from human patients. But as they started testing their system on real life data from 1998 to 2001, the system was not as accurate as expected so they decided to stop it because it failed at finding those cancer cells between 2002 and 2009.
How can we create this type of bias?
In order to avoid the automation bias from ever happening, designers and developers must understand that people do not like changes and adapt their products with them in mind. As Professor David Autor told: « People can be very slow to learn new things and adapt to change-even when the change is for the better. »
How to avoid this type of bias?
Designers should keep in mind as they create a product, as it will impact how users interact with it. There are many methods for creating an unbiased design such as designing iteratively, keeping your focus on users’ goals, following user research methodologies or using personas instead of generic users.
Examples of the Automation bias:
The idea of self-driving cars is one of the most known automation biases. Google has invested heavily in developing an autonomous vehicle that would replace human drivers. Even though they have tested their system in real life conditions, accidents involving their cars are still occurring. One example it’s when a Tesla autopilot crashed into a truck which was making its way slowly across the road at 80 kilometres per hour (50 mph).
Another major company which used this bias to make profit is Amazon after they created « Kiva » robots for moving items around warehouses. However, after launching them it became clear that the data they collected to improve efficiency could be used for other purposes like improving productivity rates in order to reduce workforce costs in places like Europe where taxes are high.
The last example comes from a photojournalist, James Nachtwey, who was an early adopter of a new robotic camera that could shoot in place while keeping a subject in frame. However after some time the system’s automation made him feel like he wasn’t carrying his responsibility as he didn’t frame his subjects properly.
Which profession use the bias?
This type of bias is more common for people working with IT or engineering fields where they are responsible for developing new products or services that impact how users interact with it. But it can also happen to designers and marketers which are also responsible for creating designs and content related to those new products/services.
Conclusion:
Automation bias is dangerous because it masks normal human behavior and makes it difficult for them to understand complex systems. It also has a negative impact for businesses or other institutions that use automation to improve their own functionalities, but this comes at the expense of employees. While companies lose money because they try to cut costs by replacing humans with technology, workers may lose their jobs too. Although there are ways in which you can avoid this type of bias such as designing iteratively or testing your product before launching it, many times we see how designers/developers choose not to do so and focus on creating flawless products instead. That’s why it is important for users of those products/services to be educated about the way they work so they don’t face disadvantages because of new technologies that were created with no regard to how people will interact with them.
Secrets de Marketeurs 