Could AI replace coxes?

AI is being used more and more in our lives, from making artwork to automating systems we use in our daily lives. So, with it becoming easier to use, could we use it to automate the roles of a coxswain, or, will the coxswain always have a place in the boat?

Having an AI cox gives one obvious advantage: steering. It would give crews an advantage on any course. An AI cox would allow for extreme precision and accuracy that no cox would be able to achieve, such as frequent small movements on the rudder which allow for steering while not compromising on balance. This ability would transfer on any course and any boat, straight courses require precision whilst accounting for any countering winds, and bendier courses require the ability to stick to a racing line which is often not clear to coxes from their perspective. However, an AI cox would also have to deal with overtaking, and on courses as busy as the tideway, this may prove tricky, especially with the sheer number of boats on the course. With this the AI cox would also have to take into account the human error that other coxes around them will have, making overtaking a tricky task.

However, one thing an AI cox may not be able to replicate is the calls. Often the calls a cox makes can win or lose a race, and if a race were to get close, this AI cox would need to know what to do. These close races are often what makes or breaks a cox, and while an AI cox might be able to replicate what a human cox would do, perhaps better, it may not be as effective. A cox is able to feed off the stroke and the crew’s mood to choose what to do, giving a sense of humility in the calls, and making the crew feel as if they are in it together. However, an AI cox would not be able to achieve that humility, simply because they aren’t human, and in order for an AI cox to be effective, they would need to be able to gain feedback in real-time from the crew, which may be hard in a race situation. Though coxing has been modernised through the use of technology allowing coxes to see the exact speed and splits of their crew in real-time, a cox still needs to gauge how the crew is feeling in the moment to assess and decide what the best course of action is.

There also begs the question of weight, an AI system would be lighter than most coxes and only accounted for in wires, monitors and hardware, hence causing less drag. But with such little space taken up by the system, I would question the need for a coxing seat at all, perhaps an AI steering system would be more sensible while still allowing space for a human cox to give calls etc.

As well as communication on the water, the AI cox would also need to be able to give feedback to the crew off the water. While an AI cox may be able to do this, it would most likely be driven by stats and data, rather than a human experience inside the boat. While stats and data are important for improvement, a cox is able to spend time with the crew outside of the boat and build relationships with them, allowing them to know what else is playing into the state of the boat. This is something that an AI cox would be completely unable to do.

Overall, an AI cox could replace coxes in all aspects of coxing, however, it would not be effective in all areas. It could be beneficial for AI to replace the steering aspect, allowing the actual cox to take control if necessary, similar to the autopilot that many cars use. However, coxes won’t be fully replaced, at least for a while, as the human nature of the cox is what makes the art of coxing unique and special.

About The Author

Lily Nguyen

See author's posts