From Military To Working On AI

drawn image of calendar icon

May 22, 2024

hand drawn writing icon

Anastassia Lauterbach

An Interview with Patrick Huston, Brigadier General (U.S. Army, retired), Member of the Board (QuantumShield), Scientific Working Group on AI (FBI), AI Task Force (American Bar Association).

Patrick Huston, Brigadier General (U.S. Army, retired), Member of the Board (QuantumShield), Scientific Working Group on AI (FBI), AI Task Force (American Bar Association)

Anastassia: Could you please walk us through your leadership journey and give some examples of those pivotal moments that have led you to where you are now?

Patrick: I grew up in California and studied engineering in New York. I joined the Army after hearing my father’s stories, and I spent 35 years in uniform. I started as a helicopter pilot but after surviving some mishaps, I decided to go to law school. I started my legal career as a prosecutor, but after 9.11, my focus shifted to what I call a targeting lawyer. I helped commanders apply the laws of war and the rules of engagement during five tours in Iraq and Afghanistan. I became a general when the Pentagon was investing heavily in autonomous weapons, a.k.a. “Killer Robots.” These systems can detect and engage targets, so there are obviously significant legal and moral issues. The Pentagon realized it needed an experienced targeting lawyer. I retired from the Pentagon in 2021, and I now live in San Francisco. I advise companies, including many tech start-ups. I also serve on the FBI’s AI Task Force and the American Bar Association’s AI Task Force.

Anastassia: You mastered several transitions in your professional career. What factors helped you to be successful?

Patrick: My mom taught me many things, including the importance of putting media headlines in context. She used to say that you’ll never read a news story about the 1,000 houses that didn’t burn down last night; you’ll only hear about the one that did. She was trying to convey that you must put individual stories into context. A house burning down doesn’t mean the entire city is on fire. I’m reminded of this when I see headlines about Tesla crashes. They grab your attention and imply that an AI systems are dangerous. But if you compare their performance to human drivers, the story is very different. The question I typically ask when considering AI and other technology solutions is whether they can outperform humans.

Anastassia: What were the most significant changes you witnessed from where you started your career to now?

Patrick: Let me address that from both the human capital perspective, and the technological angle. When I joined the military in the 1980s, women were largely excluded from key roles. This culture has profoundly changed for the better. Women can serve in every role, and have earned positions the highest levels across the military. Women are now Army Rangers, and a woman commands an active-duty Army combat division. Women lead the Navy and the Coast Guard. It was short-sighted to exclude half the population from key roles, but this progress reflects society’s evolution.

In terms of technology, the transition has been extraordinary. Generative AI is a game changer. I am an AI optimist – I believe in the power, potential and promise of AI. It can improve efficiency, streamline operations and cut costs in every industry. That said, there are risks. AI replicates human biases. It hallucinates. AI can operate fast and evolve quickly, which is part of its power, but AI can rapidly escalate and spin out of control. Some jobs will be displaced. These risks are all manageable, but C-Suite executives and board members must understand these risks and take active steps to mitigate them. Overall, AI provides powerful benefits that leaders need to tap into.

Anastassia: How did the innovation landscape change in the past two decades?

Patrick: There has been a paradigm shift regarding how technology is developed. During the last century, the military and government drove a lot of innovation. Space, nuclear, and computer technologies, even the internet and GPS, were initially created for defense use, and later migrated into the commercial and civilian spheres. The new model is the opposite. Companies often develop technology for use primarily in the commercial sector. The U.S. military will explore what’s been created in the commercial sector, and buy that technology for government use. This is an important trend. We see escalating conflicts around the globe, and I am committed to helping the U.S. and its allies promote strength and deterrence through technology. Ukraine is a good example. They are fighting for their existence, and their innovative use of technology – including drones – has been a big part of their success.

Anastassia: I am not a fan of the European AI Act. It does not solve the technology risks of AI, such as design issues due to the Deep Learning architecture, biases in datasets, use of the technology in the hands of criminals (including Deep Fakes), and calibrating development and deployment processes around the human-in-the-loop paradigm. The new framework adds substantially to the costs of the nascent European AI industry. At the same time, new risks, such as consequences of augmenting the world with synthetic data, aren’t even addressed. What is your opinion about AI regulation?

Patrick: I agree and I don’t think we should expect laws and regulations to keep pace with innovation. However, I think we have broad regulatory frameworks that balance interests and provide useful guidance. I do expect to see changes to Intellectual Property laws as a result of Generative AI.

Anastassia: What advice do you have for leaders implementing AI in their organizations?

Patrick: The number one tip I have for leveraging technology is “Human-Machine Teaming.” Humans can do some things better than machines. Humans bring judgment, common sense, empathy, and, at times, humor. But machines can do other things better than we can: rapid data computations, ingestion of massive amounts of data, or tedious repetitive tasks that would bore humans. The secret is to combine humans and machines in ways that leverage the respective strengths of each.

Anastassia: When we first met in 2018, we discussed how Google employees opposed using the company’s computer vision technologies in the defense sector. Amazon argued a different position in this dispute. Taking the war in Europe and Israel, do you think there is a shift of perspectives on how the use of AI technologies in the defense industry is broadly seen?

Patrick: Yes, AI is now mainstream. It’s top-of-mind and we’ve made a lot of progress. For example, the White House and Pentagon have published Responsible AI Principles and Strategies. One of my former clients is the U.S. Secretary of Defense Lloyd Austin, and he addressed the concerns head-on by stating that “Responsible AI is the only AI we do in the Pentagon.”

Anastassia: You are involved in quantum computing. What keeps you up at night while working with leaders in this space?

Patrick: Quantum computing has tremendous potential for improving the world. Think about its ability to improve healthcare and drug development. But there are also significant security risks. Most current security and encryption technology will become obsolete once quantum computing is readily available. Consider medical records, financial transactions, intellectual property, trade secrets, and other sensitive information. There are companies working on quantum-resilient security and encryption technology, but until they are fielded, the problem keeps me up at night.

On a related note, Deepfakes also keep me up at night. The technology is cheap, easy and readily available. It is wonderful to see it in special effects for movies. But as a lawyer, I get very concerned about deepfake evidence. Disinformation and fake news propaganda can be powerful and destabilizing, whether in the context of a local government, a nation, or the whole planet. Many companies are trying to develop deepfake detection capabilities, but it’s a cat-and-mouse game.

Anastassia: What kind of leaders can build quantum computing and security companies successfully?

Patrick: Patient, hard-working problem solvers. Like the nuclear and space races, quantum computing is a complex and lengthy endeavor. People must love tackling big, challenging problems.

Anastassia: Is this best done as individuals or by team?

Patrick: I believe in the power of teams. A culture where no one wants to let their colleagues down. Teams are the foundation of unstoppable organizations. On that note, you can’t have an effective team if you tolerate jerks. No matter how good they are at their job, jerks will do more harm than good to the team’s culture and performance. I know of a law firm in Silicon Valley that has a formal written “No Assholes” rule, which I think is fantastic!

romy and roby and the secrets of sleep book cover

Book 1

Romy & Roby And the Secrets Of Sleep.

0 Comments

Submit a Comment

Your email address will not be published. Required fields are marked *

Select your currency