Data Backup Digest

There's no arguing the fact that next-gen robots on the battlefield are a controversial issue. While they certainly have the potential to save lives, the number of risks involved – including simple malfunction to the potential for hacks – make them an unrealistic option up until now. But the United States military is hoping to change that with a new robot named Libratus – which means "balanced" in Latin – and it has peculiar origins in the world of professional poker.

A Poker-Playing Robot Joins the U.S. Military

Libratus was originally unveiled in 2017 as a project by researchers at Carnegie Mellon University. Their project, which centered on advanced game theory, saw the development of the original Libratus robot – and it was highly successful. Not only did Libratus play a game of no-limit Texas Hold 'Em against four of the top human players in the world, but it actually won.

While there is some amount of luck involved in gambling, this was no fluke. Libratus used a comprehensive database of known cards, possible combinations, common moves, and more. It even learned simple nuances associated with bluffing.

The project was so successful that it ultimately sparked the startup of a brand new company – Strategy Robot – with the clear goal of adapting Libratus for military use. A contract with the U.S. Army was finally signed in August 2018.

Assuming the Post

Libratus was put into action – via a $10 million contract – in support of a new Pentagon division known as the Defense Innovation Unit. Launched in 2015, the division encourages collaboration between the U.S. government and the professionals throughout Silicon Valley.

The software driving Libratus was originally designed for playing poker, but the fundamental concepts behind its programming – known as computational game theory – can be applied elsewhere; including military simulations and, potentially, actually use on the battlefields of the future.

But there are some challenges when it comes to deploying Libratus – or a similar robot – on an actual battlefield. As pointed out by Tuomas Sandholm, leading professor of the Libratus project, there are vast differences between war-gaming and a full-blown war.

"That opens yourself up to a lot of exploitation, because the real adversary may not play according to your assumptions." Sandholm explained.

Sandholm and his team are also researching ways to make their technology smaller and more compact. He was recently quoted as saying: ''"In some applications you need it to be miniaturized, if it's onboard something. Some platforms can't carry big computers."

The CMU professor summarized his stance on artificial intelligence and next-gen technology in the military by saying: "I think AI's going to make the world a much safer place."''

Continuing the Debate on AI-Driven Warfare

Although there are pros and cons on both sides of the debate, there is progress being made across the globe – whether you like it or not. China has a clear AI strategy as part of their national defense and Vladimir Putin, current president of Russia, has expressed support for next-gen AI on the battlefield, too.