AI Deterrence

AI technology has made significant advancements in the U.S. military. In the U.S. Navy, AI is providing a significant advantage in force-on-force engagements and the disrupting advantages of asymmetric warfare. Coupled with autonomy in logistics support, AI is poised to significantly change the battlefield of the 21st Century. Leveraging this technology will prove invaluable in the next ten years.

Therefore, we propose to introduce a new deterrence and international-relations doctrine — Artificial Intelligence Deterrence (AID). The United States can lead on this front in three key areas of AI: Autonomy, Cyber, and Economics (ACE). Within these disciplines there exists a robust ability to ensure domestic security for the United States while economically incentivizing governments to cooperate with international courtesies.

AID can be used to provide peace similar to how nuclear deterrence has provided peace since World War II. The U.S. nuclear deterrence doctrine is based on a “Triad” of air, sea, and silo delivery systems, all controlled by the Airborne Launch Control System (ALCS) — the communications link between the President and the U.S. nuclear arsenal. In reference to AID, we propose a second deterrence doctrine related to Autonomy, Cyber, and Economics (ACE).

There is currently a political shift in the United States that is driven by the return of putting the power to control our government in the hands of the People. This political and cultural movement requires an international-relations and military doctrine rooted in the moral, ethical, and honorable principles fought for and achieved on the backs of our ancestors and founding fathers. This doctrine is purely defensive — and inspires other nations to behave with the same moral, ethical, and honorable conduct the United States was once known for.

Economy

As the United States economy transforms into an AI-based economy, we will see optimizations and abundance not seen before in human history. This will result in the cost of goods — and eventually services — dropping significantly in the next ten years. The Trump administration will need to build a foundation of an AI economy that is based in innovation. Regulation must be minimal while keeping safety and AI ethics at the forefront.

Specifically, we will need to ensure that the most impressionable people (children) are not adversely affected by algorithms that cause emotional or mental harm. This will require scrutiny and accountability for some social-media companies and other companies that could potentially influence children.

Three policy areas to reach AGI

1. Meeting future energy demands. There is currently a shortfall of energy in the United States, and we must have a roadmap with robust and achievable milestones to obtain Artificial General Intelligence (AGI) and Artificial Super Intelligence (ASI).
2. Infrastructure and manufacturing. The $1.6 billion leveraged for Utah and Texas through the federal CHIPS and Science Act for semiconductor manufacturing is an example of using legislative dollars to increase capacity, access, and supply-chain capability. While this is significant, government intervention is required to ensure environmental resources are met — cooling needs for this technology transition must be balanced against agriculture, domestic consumption, and the need to preserve our environment. Without that balance, industry may do significant harm to an ecosystem or watershed if allowed to consume excessive water for a given geographical area. State or federal legislation may be required to deter over-consumption.

   A further concern is the amount and quality of training data available to AI models. Synthetic data and new approaches such as humanoid robots are some of the innovative ways we are addressing this. While quantity and quality matter, data rights must be maintained in the United States. Analysis of informed consent laws or other provisions for third-party data brokers may be needed — while AI is important, it is critical we don’t drive AI at the expense of the U.S. citizen.

3. Incentivizing innovation through policy and legislation. Certain key areas of R&D must be incentivized to realize the AGI objective. Innovative examples include Torus energy-storage technology and Nvidia’s new GPUs — most significantly Nvidia’s new chipset design, which uses Tensor Cores that accelerate matrix operations essential for deep-learning algorithms. Nvidia’s GPUs can complete training epochs for neural networks up to 28.4 times faster than 8-core CPUs, significantly accelerating the development cycle for AI technologies. After Nvidia introduced Compute Unified Device Architecture (CUDA), a broader range of industries and researchers gained access to powerful computing resources for AI development. From 3D stacking of chipsets to the innovation AI brings to manufacturing, innovation must be identified and effectively leveraged to keep pace with emerging global economies.

Currency & Sanctions

The United States’ economic influence over the span of 20 years has waned for several reasons. The misuse of sanctions has notably contributed: U.S. currency on the world stage has lost approximately 15% over the last 30 years. While 15% is a conservative number, the implications for projecting power through sanctions are at risk. There is currently a move by China and Russia to return to the gold standard through BRICS, which would effectively remove the dollar as the dominant currency and suspend the United States’ ability to use sanctions to persuade behavior.

Iran & Afghanistan

Most notably is the abuse — or absence — of sanctions on countries that have behaved badly. Iran is a clear example. The Biden administration’s policy significantly influenced the respect of the dollar by allowing Iran and other terrorist organizations to receive money when we knew they were going to spend it for nefarious purposes.

The Biden administration also colluded with the United Nations to send $2.9 billion in U.S. currency to Afghanistan — $1.8 billion in 2022 and $1.1 billion in 2023, per a report from the Special Inspector General for Afghanistan Reconstruction (SIGAR). These actions, coupled with the blatant disregard for the Taliban failing to meet any of the provisions in the Doha Agreement, have disrupted international trade and funded an organization we know does not or cannot meet its obligations.

While the Biden administration has done much harm to how the United States conducts its financial relationships with the international community, there are opportunities to implement provisions and safeguards to ensure this behavior does not continue with another administration. Specific reporting, auditing, and accountability need to be baked into our financial systems in order to keep our Executive Branch in check and ensure those systems become more transparent.

Autonomy

While “autonomy” is an economic driver, it is also a U.S. military driver. With the effort and accomplishments the U.S. military has made, it is poised to significantly replace human combatants and substantially increase our ability to deter peer competitors. AI and autonomy in the military is a force multiplier and will change the landscape of how we conduct war. This comes with dangerous and powerful responsibility for our representative government — sending our children to war will not be a significant motivator as it has been in the past.

In the DoD, significant efforts are being made to advance autonomy in weapon systems and aircraft. This effort, while important, should have the appropriate amount of oversight from Congress. Industry compartmentalization, security, and ethical warfare must be maintained while these technologies are implemented.

Implementing AI in non-lethal Command and Control (C2) operations should be paralleled with implementing and optimizing algorithms in autonomous vehicles to obtain data-fusion techniques at the necessity of operational tempo. Engineers and scientists will need access to critical AI education in Large Language Models (LLMs) and eventually Large Quantitative Models (LQMs). That will translate to unclassified training with organizations like Defense Acquisition University (DAU) and classified training in how to leverage and collaborate with federal and industry labs to make program-level integration of AI possible.

Cyber

Cyberwarfare, cyber security, and cyber compliance will be some of the main areas of concern in the near future. Most significantly is the United States’ transition from our current encryption technology to a technology that ensures adversary AI is not capable of breaking it. The National Institute of Standards and Technology (NIST) has recently introduced measures for Post-Quantum Cryptography (PQC), which provides means to deter the intrusive advantage of nefarious quantum computing — anticipated to increasingly become a threat as we approach AGI.

References

1. Rep. Tim Burchett — “Your tax dollars are going to the Taliban.”
2. Brookings — Around the halls: implications of the U.S.-Taliban agreement.

Under construction — more to come.