Exploring AI and the Future of Defense: TMI23 in Review


Yesterday at HyperWerx, national security, government, and defense leaders from around the world gathered to discuss and explore real-world applications of artificial intelligence and other emerging technologies today. SparkCognition Government Systems’ “Time Machine Interactive: AI and the Future of Defense” (TMI23) assembled a powerhouse slate of panel discussions, expert presentations, and live demonstrations that could be characterized as eye-opening, gut-checking, action-oriented, and incredibly thought-provoking.

A heady lineup of distinguished guests took on the vast topic of military readiness through the lens of AI, unpacking its current and evolving implications for international relations, economies, ethical questions, and government and commercial partnerships.

In his welcoming remarks to the attendees, SparkCognition Founder and CEO Amir Husain framed the current moment in unequivocal terms: that the AI disruption that futurists have been talking about for decades is not only here—it is taking on an “emergent” quality. He clarified: “You might say “emergent” doesn’t fit here, but I’ll explain to you how it does. The AI disruption is becoming emergent. It was not obvious; it was not constructed. It is becoming emergent.” 

Husain acknowledged that any great technological aspiration would attract its fair share of detractors and naysayers. But, he said, anyone looking at the facts—what AI is already doing today and the blistering rate of advancement in machine learning, generative AI, etc.—and still argue that AI technology will not fundamentally and rapidly change the future of defense as well as how government, industry, and society operate needs to recalibrate their opinion, lest they be left behind. 

“We’ve just gone on and done it.”

One by one, he refuted assumptions that AI is no more than an extension of what we already know and do with statistics, that it doesn’t scale well, that it can’t outperform domain experts, or that it’s not powerful enough to disrupt the biggest players in global relations or industry. 

In answer, Husain sketched the story of how and why SparkCognition, SGS, and SkyGrid formed. Because the world needs change, the world needs change-makers, and AI is the disruptive type of technology that will enable the change we want to make:

Now, of course, I am an unapologetic believer in artificial intelligence, and I think it’s the most important thing that’ll happen during our lifetimes and probably for a long while thereafter. So with that caveat, let me tell you what my view of these things is. When we started SparkCognition, I encountered all of the criticisms that I just shared with you. 

Our dream with SparkCognition was to apply AI to the physical: to create a marriage between atoms and bits to imbue physical systems with intelligence. We weren’t satisfied just recommending the next Netflix show or optimizing the click rate on an ad. We wanted to bring artificial intelligence to global infrastructure, which is a $100T problem.

And unless we improve this global infrastructure by making it safer, by making it easier for people that are less experienced to use, and by reducing the carbon footprint of this infrastructure, then we are in a world of hurt. Because what do we do? Either bear the consequences or rip and replace $100T worth of stuff. Not doable. 

So the only answer to infrastructure is going to be artificial intelligence—add the software that upgrades the hardware and makes it run better. That’s the net-net. 

SparkCognition was conceived as the first industrial AI company that was looking to build an operating system for global infrastructure, which is a huge vision. But over the years and over the months, that vision took hold, and a group of partners emerged that helped us along with this vision. 

With General John Allen, I worked on the idea of the application of AI in the field of battle, and from that, a whole body of work emerged this idea of HyperWar. And at the heart of that idea was the notion that the ‘Observe, Orient, Decide, Act’ (OODA) loop […] this whole notion of perception, decision, action, that that could be compressed down to zero with AI-based systems integration, data integration, automatic targeting, and also kinetic control. So you would be in an environment where things would be happening faster than a human could respond. And then we projected that out and we said, what does HyperWar look like? What does it mean for teeth-to-tail ratios in militaries? What does it mean for the Western training advantage? On manned platforms today, our fighter pilots are the best because they fly the best machines and they get the most training, 250, 225 hours a year. What happens when, with autonomy, an opponent does not require 220 hours to best a pilot that has put in that kind of effort in flying planes? 

With Boeing, then we established SkyGrid. The idea there was clearly the future of defense is not thousands of manned aircraft in the sky. The future is millions of aircraft in the sky that are autonomously working and, and these are things that you might not even conventionally call aircraft, but will do things in our society economically, militarily, and logistically that will change the world.

[We worked] with my friend Michael Brasseur, who’s sitting here and is a real pioneer as the commander of Taskforce 59, along with Michael Stewart, his partner on this project. They established the Navy’s first autonomy operation through TaskForce 59. And when they did that, we partnered with them to create this idea of the “digital ocean” working with NATO and then also deployed these capabilities and demonstrated real firsts at IMX. And then of course, there is all the work that the SGS team are doing on the JADC2, in particularly re-imagining this with AI sensors. 

So my point here is that for all that criticism over the last ten or 12 years, like: “You’re not even as good as a vibration analyst,”; “Your AI will never be able to solve a domain focus problem,”; “You will never be able to achieve autonomy. Nobody will let you.”—we’ve just gone on and done it.

And that is part of the spirit of Time Machine Interactive. One of the things that I really want our friends to take away from [TMI23] is: “Damn the torpedoes.” If you want to build the future, every large organization will somehow, deliberately, or otherwise put some kind of obstruction in front of you. 

People don’t generally like change, but the change-makers have to bring about that change. And so no matter how hard and no matter how small we were when we started, these are just facts. These are factual things that we’ve done. We’ve deployed now on hundreds of thousands of cameras in 16 countries. We have 130,000 people using SkyGrid software. We have gigawatts of energy under management. Over the last three years, our AI software has shipped about two to 3 billion barrels of oil. We are making a small difference toward that vision. 

So when you leave here, hopefully, my other dream and desire is for you all to leave here as friends because you are brilliant people, extremely accomplished people with great connectivity, with amazing organizations. And imagine just discussing, talking, getting to know each other, and then figuring out what we might all do as a group: to be able to do it together. There’s going to be hopefully a lot of collaboration that comes out of this.”

By all accounts, SparkCognition Government Systems’ “Time Machine Interactive: AI and the Future of Defense” was a resounding success. In the coming days and weeks, watch this space and sparkgov.ai for more recaps of all the sessions. We can’t wait to share more of what we learned together.

We want to thank all who attended, our staff who worked tirelessly to coordinate our seventh(!) annual TMI event, and our tremendous sponsors: Boeing, Premise, SkyGrid, and Silicon Valley Defense Group

Latest blogs

SparkCognition is committed to compliance with applicable privacy laws, including GDPR, and we provide related assurances in our contractual commitments. Click here to review our Cookie & Privacy Policy.