Skip to content
Daisie RegisterAugust 31 20215 min read

The Problem with One-Size-Fits-All AI

What can artificial intelligence do for your agency? The answer isn’t all that simple: it depends on a number of factors, like how your agency operates today and what your goals are for the future. AI solutions, therefore, are most effective when they’re customized to meet those specific needs.

Few organizations know this better than the Department of Defense (DoD). As the largest federal agency, DoD has made several strides to prioritize and implement AI across agencies in a way that contributes to the enhancement of national security.

In this third and final episode of Machine Momentum, Iron Bow Technologies’ Jim Smid and James Ebeler discuss the important work DoD is doing to implement customizable and scalable AI solutions.

 

The DoD is spearheading a lot of AI-focused initiatives through the Joint Artificial Intelligence Center (JAIC), how has this center of excellence accelerated AI adoption across the organization?

The DoD has done a great job at enabling the JAIC to ensure all parts of the DoD is moving in the right direction as it pertains to AI. Having one, centralized organization that provides guidance, processes and outcomes allows for the service to really focus on specific mission requirements. Specifically, one thing JAIC has done extremely well is having a Terms of Reference to ensure every branch has the same definitions as they move forward. This allows them to move out quickly and with a sense of urgency. And so although the JAIC has been central to AI for the DoD, each service has its own AI initiatives and is driving its own AI adoption across most mission areas as it pertains to each service, staying within the limits of the JAIC.

 

There’s out of the box AI and then there’s customizable AI. Can you speak to how the DoD is using customization and why that’s important as it pertains to some of the solutions they’re looking for?

I would say six out of 10 times out of the box AI solutions will work for most customers to use its foundational capability. However, it’s definitely not a one-size-fits-all technology or solution. AI can be best employed to make targeting quicker, to make logistics more timely and precise and to improve analysis and tactical operations. Additionally, having the flexibility to change and optimize those capabilities as we go down the AI path is key.

 

Can you expand on where you’re seeing AI in the DoD and where people are seeing the most success?

Currently, there are numerous DoD agencies employing AI in their missions and they’ve all been successful. One is Project Maven for the Air Force. In fiscal year 2022, the Air Force will be looking for ways to leverage AI and ML to get ahead of its maintenance demands which the service believes will create savings that can then be reallocated.

Moving forward, working with industry to be able to customize all solution sets will also be key. It will be important to keep an eye on how well they can leverage AI as it comes to targeting. We’ll never take the human out of the middle of these operations, but leveraging AI to ensure we’re on target is going to be key as we go down the AI path.

 

As we start to “take the human out,” it begs the question around trusting the AI and trusting what you’ve created. Can you talk about the importance of that human element and the SMEs that are developing your AI?

As it pertains to the trust of the AI, letting the human let go and actually having the machine make a decision for them, it’s crucial that the human is comfortable with the code or with the platform they are leveraging in order to do that, so being part of the buildup and development is critical to fostering that comfort level. It will come down to all the data points that feed into the AI platform. It’s an iterative process to go from developing those algorithms, analyzing the data over and over, and getting to the place where it has a dependable degree of preciseness.

 

Last podcast we talked about security, can you talk about how the DoD is using AI in security and trying to assist humans with their security processes?

AI is really posed to transform every industry and is expected to impact all corners of the DoD: operations, training, sustainment, force protection, you name it. With the application of AI for Defense, we’re going to have the opportunity to improve security, support and protection of our U.S. service members, safeguard our citizens, defend our allies and improve the affordability and speed of our operations.

As it pertains to IT security specifically, AI can and is in the beginning of playing a role in cyber security through our supply chain risk management. AI is also going to greatly enable the implementation of zero trust architecture. We’re already seeing a lot of software companies building some sort of an AI component into their software platforms which will allow for quicker decision making as we move forward. And once fully implemented, AI will help the DoD adopt a proactive risk posture as opposed to our historical reactive approach.

 

Any advice or takeaways that you would leave us, the DoD or other agencies with as they are tailoring AI to their organizations’ needs?

One thing organizations need to think about is not focusing on AI in a vacuum, there are plenty of resources out here to help you so engage with industry like Iron Bow, academia and our international partners. Always keep focus on the unit mission and consider how best to apply AI to those tasks. AI is best incorporated as a supplemental element across the organization’s overall strategy to improve operational effectiveness and help make decisions quicker and more accurate.

Finally, consider how AI can change an organization’s needs, how it can reduce some current needs but create new ones. As we go down the AI path, some resources may be able to be reallocated as AI automates some processes so focus on what current needs are but keep in mind that with leveraging automation, AI is going to create new opportunities for humans down the line.

This is the third and final podcast in this Machine Momentum series. The first can be found here and the second can be found here. For more information on how Iron Bow is utilizing AI to assist DoD organizations, click here.

COMMENTS