How to think about military AI, not what to think.
Midjourney prompt: How to think about military AI
I'm an evangelist, someone who passionately believes that they have seen the future. My future belief is that artificial intelligence, automation, machine learning presents an opportunity to enhance human ingenuity far beyond any previous technology.
It also has the potential to cause great harm.
I am sometimes accused as someone who thinks differently, and I am not always sure that is meant as a compliment. Yet it is in this perspective, of a society grappling with major changes and upheaval through automation, that I urge military leaders to consider HOW they think about AI, and to move away from a narrative that is mostly focussed, if at all, on WHAT to think.
What do I mean by this difference? When the US and UK adopted a doctrinal approach to warfare in the 1980s, they adopted an approach that enables soldiers to think about their situations and decisions rather than follow a simple drill or routine. This change was resisted by generations of military officers, shaped by deterring Soviet aggression in Europe, more used to prepared positions with a sequence of triggers and responses than independent action. I experienced this resistance.
We began to encourage leaders to learn how to think about any problem rather than what answer to provide to a specific problem.
Yet technology adoption has been blighted by a preference to deconstruct innovative technology rather than understand the opportunity, especially in the UK.
The impact of the machine gun in 1914 led to the creation of The Machine Gun Corps in 1915. Their role was to understand the technical aspect of machine guns and to deploy them more effectively. Eventually, in 1922, The Machine Gun Corps was disbanded, and its role and purpose embedded intrinsically into infantry brigades. We took seven years to truly integrate this essential weapon.
The pattern was repeated with the tank, where the Heavy Machine Gun Corps owned the technical adoption and maintenance of the tank as mobile machine gun posts. Thinking on how to deploy tanks came later. It is probable that the decision to limit Germany's physical access to armoured vehicles accelerated its thinking on how to deploy tanks in new ways rather than languish in admiring their mechanics.
In both cases, the emphasis was on a specialist group understanding a specialist equipment and eventually implementing that equipment on the battlefield. We deconstructed the system, understood it at a systems level, and eventually became comfortable with the technology over many years before integration.
We are too often continuing this archaic approach with automation and AI. The utility of AI is explored by specialist units and procurement parts of defence, usually innovation or experiment teams, and is seldom experienced by or involves troops struggling with genuine issues. Bolder military elements are attempting to get AI into the hands of users, but their foresight is often hindered by both access to soldiers and support for development.
Like drones and cyber before, the opportunities to explore and experience automation on a general military exercise are few. Automation exercises are designed, established, and run purely to explore the AI system rather than examine its integration and exploitation with other military systems. Large military exercises are conducted without any automation.
This training focus may mean that large troop movements can be conducted smoothly across exercise plains and areas. These exercises may validate that current equipment continues to operate as expected, or that commanders are able to deploy tactical manoeuvres. It avoids training exercise serials and schedules being disrupted by automation.
Yet current exercises are not preparing the next generation of leaders for warfare in the intelligent age.
Leaders and those led, at every level, need to see what is possible with automation and to explore and exceed its limitations. It is only by using an equipment that empowers soldiers to learn about its true utility, and it is only through experience with innovative technologies that their disruptive nature can be explored or developed. Technology elements may not be ready for full deployment, but if commanders are not thinking about how they could use AI, even if theoretically, they will miss the opportunity to shape how they will use AI.
Our soldiers are often disruptive, and military leaders need to fire their enthusiasm for how AI will disrupt their core business.
Today, it is often too easy for a military leader to deny that they have technological understanding unless they are in a technical branch. There remains a certain badge of honour to declare that they know nothing of mobile phones, or do not use applications, or have no time for information technology. Defence appears unique in its desire to spend so much of its budget on technology and so many of its leaders to flout it.
Some remain convinced that basic navigational skills are more useful than GPS or that fitness trumps technology. This misses the fundamental point that future successful military leaders will be both fit and understand technology, they will still understand the traditional and enhance it with technology.
This is the truth behind automation – it enhances human ingenuity and defence is not separate from this disruption.
Military leaders who eschew this automated enhancement will not be championed but be defeated.
Therefore, I propose that:
Current military academies and training centres seek to implement and adopt automated tools and practices in every aspect of training.
We need to encourage current and future leader that understanding and exploiting AI is at the very heart of their thinking about warfare.
I have said before that doctrine needs to be rewritten with automation in every element, questioning how it changes our approach to conflict.
We need to develop insatiable curiosity about the potential of technology within all our military leaders, especially the fighting arms.
We must generate a cohort of leaders who will test and explore the greater potential of technology rather than ignore or abhor it.
Future leaders are curious, not curios. They need our help to foster and fuel their curiosity.
