This morning, I read an article from The Wall Street Journal titled “Israel Was Prepared for a Different War.” The article focused on how Israel recently prepared for a different kind of war, but the attack was executed differently. I think the opening example of the article illustrates the gist of the title of the article beautifully:

“Israel spent three years building a 40-mile-long, state-of-the-art, high-tech security barrier along the Gaza strip, with radar and sensors designed to detect furtive incursions by Palestinians, bent on carrying out covert attacks in Israel.

On Saturday, Hamas used bulldozers and other rudimentary means to punch through the 20-foot-high fence and flooded men through the gaps…..”

Humanly, it is not possible for us to accurately foretell all the possible ways something can go wrong. Fortunately, AI can help do exactly that!

We have the tendency to associate technology with the technology hardware itself. This becomes much more prominent when building hi-tech armed forces capabilities since the prowess of an armed force has traditionally been displayed through arrays of weapons.

And hence, it is definitely critical that those technologies become hi-tech. I don’t think that Israel investing in ensuring that the fence was hi-tech was a mistake. It was an imperative. I think the article tends to bring forward that Israel was not that prepared for a different war. The armed forces of a country with Israel’s history need to prepare for all kinds of war. The issue lies in the words “all kinds”. Israel should also have prepared for the type of invasion that ended up happening.

Before I proceed further, this article does not aim to provide a political opinion. I believe in political views and do not believe in neutrality. If something wrong is happening, a man with no opinion is complicit. But this is not the post for expressing that opinion.

One of the thematic messages of the book “The Art of War” is that- “Your Mind is your Best Weapon.” Though you need both the resources (ground troops, assets, and equipment) and strategy, almost all wars are won due to strategy. Leading war schools across the globe tend to teach lessons from all the wars and battles that have happened across the globe to highlight “lessons learned.” As this incident highlights, sometimes the lesson learned may be hidden in an event yet to occur. And this event, which has not yet happened, can be “generated” by Generative AI.

When we got bored with poems written by Generative AI, we became engaged with fake citations generated by AI. Have you read some of them? They were so well articulated; no wonder the poor attorney did not validate them before using them. But keeping that “magic” of AI hallucination example apart, can you not build an LLM model to learn from all the wars, generate all possible scenarios, and use the “magic” to hallucinate and generate unimaginable scenarios? In this case, the hallucination will not get anyone in trouble but will instead provide another scenario to consider to be prepared.

And if you are still brooding to answer my question, let me help. The answer is yes; you can build an LLM that can:

  • “Learn” from every battle and war that has ever happened and then extrapolate them to scenarios based on your specific geopolitical situation.
  • Generate new scenarios, every possible one, and reconcile it with your existing resources to highlight gaps. Flags all scenarios where the resource gap is high enough.

As the world becomes more chaotic and unpredictable, the limitations of our capability to learn everything that can go wrong become evident. And this is where we can leverage the true power of AI technologies, like Generative AI. Because poems and pictures can wait, but human lives are precious.


One response to “Generative AI and Warfare”

  1. Chaos Theory and AI: Replacing Emotions with Equations (Part II of II) – Designed Analytics BLOG Avatar

    […] the realm of dynamical systems than pure chaos theory. A version of the algorithm I suggested in this article, can help predict the system’s behavior. The approach will be similar to how we forecast […]

    Like

Leave a comment