Mitigating Predictive Processing Errors

Andy Clark’s book “The Experience Machine: How Our Minds Predict and Shape Reality” was just published in May. The science he discusses in the book is not new, but the methods used to understand our brain are relatively new. This begs the question: Are we basing our safety training on an older model of the human brain?

That older model is an orderly cause-and-effect/stimulus-and-response system. Essentially, if we train to respond appropriately to stimuli, we think we are all set. We imagine our brain patiently awaiting stimuli and sensory information so it can send out the appropriate responses. Much of the older model was derived from brain autopsies. But as Tan Le states in her book “The NeuroGeneration,” “Dead brains don’t talk.”

The Reality
In reality, the human brain does not passively wait for information to which to react. It is constantly predicting our next move and creating an action plan based on its suspected prediction. Think of watering your plants; if you wait to see if your plant is wilting, then your action is too late. It is better to predict that your plant will need watering at regular intervals. Your prediction as to when to water the plant is based on experience. The brain’s job is to water all the things in your life before your life wilts and ends. By the way, your predictive brain is also the reason you can’t tickle yourself.

Training and preparedness are both good things. The human brain, however, is in the business of predicting risk. Since the brain acts first and senses second, we need to provide an internal model for our prediction to be accurate. Our intuition and predictions are best if we actually believe in the possible outcomes. If we don’t believe an outcome is possible, our brains are unlikely to produce a safety prediction to avoid it. If we have no experience from which to predict, then we are said to be experientially blind (for more on this topic, watch Lisa Feldman Barrett’s TED Talk at www.youtube.com/watch?v=0gks6ceq4eQ). In other words, if we have not created a circuit to recognize the hazard, our brain will be unable to predict the right response. We create new circuits from learning and experience. When we touch our first hot pan as a child, we create an experience that slows our action when we are near another suspected hot pan. Our subconscious creates a space between reaching for and touching that pan.

Changing the Safety Game
So, how does this change anything?

Here’s how: If we can get our neurons that recognize what a bad outcome looks like to wire together, we will have information to change the response to our brain’s prediction. For example, if on a hot day you performed a brief mental rehearsal with your crew on a response to a man down, the response played out between crew members will fire neurons that wire together. Within three to five minutes, you should be able to sequence what a response to the fallen worker looks like and have two runs at it, including a debrief.

Most likely, the first-aid training has been done, but sequencing a response – and experiencing the stress of that response – has not been practiced. We practice the skill but not for the event. The reason we don’t consider practicing for a bad event is most likely optimism bias (for more, check out Tali Sharot’s book “The Optimism Bias: A Tour of the Irrationally Positive Brain”). It is good that we are optimistic; however, it means that we don’t like to consider what we would do in an unfortunate situation. For example, even if you have been trained in CPR, it is unlikely that you have ever sequenced what you would do if, when you returned home today, your significant other collapsed in front of you and you couldn’t find their heartbeat. The typical response is to call 911 and start CPR. But what is quality CPR, and what will make a difference for your loved one? Dealing with the stress of that situation is rarely if ever practiced – even though approximately 1,000 cardiac arrests occur outside a hospital environment every day in the U.S., with 70% of those occurring in the home. The frequency of cardiac arrest suggests we should be ready for it. But if it were to happen to your spouse today, would it be your first dry run? You are experientially blind to that ever happening because you are optimistic that it won’t – until it does.

The Job Brief as an Engagement Tool
In our industry, if we can learn to use the job brief as a tool to engage crew members and get synchronized as a group – instead of only having one person read the brief, which often sounds quite similar to the one read the day before – we might find that we are more prepared for the risks that potentially follow. Immediately after the job brief, if everyone identifies their job role, their immediate plan of action and any concerns they have, it will create a space in time to recognize risk that may have been missed. When you hear a crew member say, for example, “I’m a little concerned about that tie wire on the corner pole when we raise the primary 8 feet,” that may create an action that needs to be performed (e.g., “Let’s retie the insulator, or better yet, let’s hold the phase with a grip.”).

Our crews often have the knowledge necessary to work safely. In fact, it is seldom a lack of knowledge that causes incidents. It is more likely an inability to apply the knowledge they have. Creating a connected crew through a brief discussion about roles and concerns can change the crew’s cumulative predictive processing.

In closing, we must create an environment that allows time for discussion and engagement, which in turn allows for ideas to be called up from our subconscious. Controlling risk requires us to predict risk. Conscious predictions that we write down on the job brief are not any more accurate than subconscious predictions we gain from each other.

We are smarter together.

About the Author: Bill Martin, CUSP, NRP, RN, DIMM, is the president and CEO of Think Tank Project LLC (www.thinkprojectllc.com). He has held previous roles as a lineman, line supervisor, project manager and safety director.