The onset of AI in the workplace raises key questions: How can we use artificial intelligence to help us constantly get better at our jobs, rather than eliminating them? Provided AI will automate the mundane, can it also train staff to rise above those tasks as they disappear?
We don’t view artificial intelligence as a threat that replaces us, but as a multiplier on our personal skills. We realise that not all intelligence is artificial. Dispatchers or transport planners bring a complimentary problem solving approach, and different solution options. An activity as pervasive and diverse as trucking needs all this firepower.
A synergy between software and staff offers a richer set of tools for value creation. It also leads to greater job satisfaction and promotion into higher paying roles for the planner staff.
The best use of AI in enterprise software is to guide workers toward doing their jobs more effectively while they’re doing it. It is a coaching resource, where they retain agency to make meaningful decisions. Within TNX, a transport planner is being guided by effectively limitless memorisation and computation, but the human is the final arbiter.
By allowing meaningful choices for the users, TNX supports development of personal style, expertise, and ultimately better creative outcomes. As these are incorporated into the data TNX can learn from, everyone benefits.
On the other hand, there are substantial and destructive biases in human decision making. Because they are subconscious, they present as bias disguised as wisdom. TNX adds value in part by helping dispatchers free themselves of biases in how they make hundreds of small transport decisions each day.
The core of how AI works with staff is in framing decisions for the user in a way that counteracts their bias and but still allows for meaningful choice, style, and creativity. In a software context, this can be expressed in many ways including sortation of options, tagging or text descriptions, displaying figures in absolute vs. relative values, or just how a screen is laid out. The example below comes from the TNX app, and notice how we use a small shake of the call to action button to draw the user's attention to where they should look.
At TNX, this translates into a two-part focus. First, the "AI in the middle" needs to produce excellent suggested actions. Then, the frontend software needs to present options to the user in a way that exposes the true tradeoffs but allows for individual style and agency. Our vision of AI for trucking puts as much emphasis on the frontend, where staff interact with the software, as the backend where the advanced computation takes place.
To see how this will feel, logistics leaders should consider the appearance of AI in medicine. Logistics decisions have a similar mix of messy physicality, many data sources, and emphasis on experience. Now look at AI's inroads to x-ray reading, case-history assessing, and even diagnosis. Doctors are giving up tedious and exacting tasks to their AI assistants and emerging happier and more effective as a result. The role of the transport planner (like the doctor) is redefined, not abolished.