As artificial intelligence (AI) continues to grow in importance for modern military operations, the US must build the necessary infrastructure needed to keep pace with adversarial AI capabilities.
An aging power grid and lack of on-shore manufacturing capacity for semiconductors are areas of concern. China is investing heavily in AI, in both defense and civilian fields. China has also invested heavily in power capacity to meet the energy needs of AI research, development and implementation.
To match this, the administration in July released its AI Action Plan, which calls for major investments in infrastructure and manufacturing. Meeting these goals will require taking advantage of private research and development for public goals, said Steve Orrin, Federal Security Research Director and a Senior Principal Engineer with Intel.
“The way to think about chip innovation for public sector is understanding that public sector writ large is almost a macro of the broader private sector industries,” said Orrin. “Across the federal government and public sector ecosystem, with some exceptions, you’ll find almost every kind of use case with much of the same usages and requirements that you find across multiple industries in the private sector: logistics and supply chain management, facilities operations and manufacturing, healthcare, and finance.”
The military use case for AI
While military officials stress the importance of humans being in the decision-making loop, AI offers a number of uses that do not require constant human oversight. For example, AI can sort through reams of reconnaissance data to identify patterns or changes in areas that are under surveillance, and it can do so much faster than human analysts. That information can then be shared with analysts to evaluate what the patterns or changes mean, informing intelligence reporting and decision making.
AI also offers advantages in areas such as predictive maintenance, logistics and operations. It can be used to detect potential flaws or breakdowns in equipment so they can be fixed before equipment goes out of service. For logisticians, AI can sift through data on the environment and assets available in an area, allowing for better analysis and decision making on resource allocation. In operations, AI can be used to identify unknown objects, track troops and equipment in the field, and even deploy autonomous vehicles. All of those applications need computing power that is accessible in remote or contested environments, which means cloud-based solutions are key.
“Being able to take enterprise-level capabilities and move them into edge and theater operations where you don’t necessarily have large-scale cloud infrastructure or other network access means you have to be more self-contained, more mobile,” Orrin said. “It’s about innovations that address specific mission needs.”
Rebuilding the grid
AI is power hungry in all stages, from research to development to deployment. Meeting the power demands requires a grid capable of supplying the necessary resources, and the grid in the US is aging and prone to disruption from natural disasters, extreme weather and breakdowns. Increased demand from electrification of vehicles and more personal devices that need charging puts additional strain on the grid as well.
The AI Action Plan calls for four policy actions to ensure the grid can hold up to the demands of AI. They are:
Stabilize the grid of today as much as possible.
Optimize existing grid resources as much as possible.
Prioritize the interconnection of reliable, dispatchable power sources as quickly as possible and embrace new energy generation sources at the technological frontier (e.g., enhanced geothermal, nuclear fission, and nuclear fusion).
Create a strategic blueprint for navigating the complex energy landscape of the 21st century.
Optimizing how AI is being used can also help ensure technology and infrastructure resources are sufficient. Being able to scale AI both horizontally and vertically helps ensure the right kind of computing infrastructure is in place, because there are different size, weight and power requirements for different tasks.
“Maybe it’s doing federated learning because there’s too much data to put it all in one place and it’s all from different sensors,” said Orrin. “There’s actually benefits to pushing that compute closer to where the data is being generated and doing federated learning out at the edge. “At the heart of why Intel is a key player in this is understanding that it’s not a one size fits all approach from a compute perspective but providing that right compute to the needs of the various places in the horizontal scale.”
The manufacturing base
Semiconductors are a critical component to developing AI, which means home-grown manufacturing of them can be a safeguard against supply chain disruptions and restrictions on sales from other countries. The AI Action Plan calls for investment in manufacturing capacity, with specific callouts for streamlining manufacturing regulations and reviewing grant and research programs to ensure they are adapting AI technology to speed up production of semiconductors.
Orrin said that ultimately, much of the work has already been done in the private sector. But for the Department of Defense and the government at large to meet its AI goals, it will need to adapt technologies that are already developed instead of reinventing the wheel.
“When we talk about chip innovation specific for the public sector, it’s this notion of taking private sector technology solutions and capabilities and federalizing them for the specific needs of the US government,” said Orrin. “There’s a lot of interplay there and, similarly, when we develop technologies for the public sector and for federal missions, oftentimes you find opportunities for commercializing those technologies to address a broader industry requirement.”
Intel semiconductors are in every aspect of the ecosystem, including cloud computing, data centers, client systems, and edge processing nodes. That gives the company expertise across the spectrum, allowing it to identify and implement the right solutions for whatever challenges may be faced.
“This holistic view of the requirements enables us to help the government adopt the right technology for their mission,” said Orrin. “That comes to the heart of what we do. What the government needs is never one-size-fits-all when it comes to solving public-sector requirements.”