Six years ago, the USS McCain collided with a Liberian oil tanker in the South China Sea, killing 10 sailors and injuring five others. Among the contributing factors cited by investigators: crew members had not been properly trained on a new software system that replaced analog throttle controls.
For Andrew Powell, CEO of Ethos (formerly Learn to Win), the McCain incident represents exactly the kind of preventable tragedy his company aims to eliminate.
“The root cause of that problem, we think Ethos is a perfect solution for,” Powell said. “There was no system that the Navy had to flag this key mismatch in operational readiness.”
Powell’s solution is what he calls the first “human readiness platform” for defense: a system that continuously tracks what military personnel need to know, measures what they actually know, and automatically delivers personalized training to close any gaps.
It’s a departure from traditional military education, which Powell argues moves too slowly for modern warfare’s rapidly evolving demands.
The company recently rebranded from Learn to Win to Ethos, capturing what Powell described as “the fundamental character and practices and culture of an organization” that determines success or failure in military operations.
Are you building game-changing tech for defense? 💪
We’re always looking to feature innovators pushing the boundaries in autonomy, AI, energy, and beyond. If your company belongs in the Defense Disruptors spotlight, get in touch with us here.
Training at the Speed of Software
Powell’s critique of current military training extends beyond tragic accidents to fundamental questions about readiness in an era of rapid technological change.
“If you look at drone tactics and counter-drone tactics, those things fundamentally change every few months on the battlefield,” he said. “How do you keep pace with that change in a legacy process that might take two to three years to define what doctrine should be?”
The answer, according to Powell, lies in what he calls “AI-defined” and “software-enabled” approaches to training that can adapt as quickly as battlefield conditions change.
Traditional military doctrine development, he noted, still relies heavily on “getting a lot of people together in a room, they discuss something, they write up a long Word document that then gets propagated out to the armed forces.”
Ethos aims to replace that process with continuous, automated updates that track changes in equipment, tactics, and procedures, then immediately trigger relevant training for affected personnel.
Using the McCain example, Powell explained how his system would work: “A change to the throttle system would be logged in our system. That would then trigger a lesson that goes out to everybody that’s associated with operating that system.”
Proving the Concept
The company’s approach has shown measurable results in initial military applications. Powell cited work with Air Education and Training Command, where Ethos helped revolutionize a nine-month introductory course that had the highest failure rate of any Air Force program.
“In the first year of using our products, they actually cut their failure rate in half from their historical baseline,” Powell claimed. “That led to, by their estimates, about $2.3 million in savings, just in the per diem associated with housing and feeding and continuing to pay salaries for people who would end up failing the course.”
The platform works by giving instructors detailed visibility into individual student progress and automatically personalizing learning experiences.
“This person’s struggling with these three topics; they need some additional instruction,” Powell explained as an example of the system’s feedback.
Beyond initial training, Ethos is being applied to operational settings. Powell described work with an Air Force maintenance organization where the platform creates “a continuous feedback loop between performance data around how the maintenance shop is performing” and targeted training interventions.
The system can identify patterns in maintenance logs — repeated misdiagnoses of particular problems, for instance — and automatically deliver training to address those specific knowledge gaps.
“Rather than taking a one-size-fits-all approach, we think AI finally enables us to be hyper-personalized about individuals’ competencies,” Powell said.
The Sports Team Model
Powell frequently draws analogies to professional sports, where Ethos also works with NFL and college football teams. The parallel, he argues, is instructive for military applications.
“An NFL team plays a game on a Sunday. They then review all the game film, and they say, hey, on this particular play, this wide receiver made this mistake, here’s the fix for the next time,” Powell explained. “They’ll build in our platform the right intervention for that player so that then the next Sunday, when they go out and compete, they’re not making that same mistake.”
This model of continuous performance analysis and targeted intervention, Powell contends, should replace the military’s traditional approach of extended initial training followed by periodic refresher courses.
“We think this very adaptive, agile approach to training is much more effective than going to school for a year or two years, you’re then certified, and then you’re kind of just off to do your job.”
Navigating Security Requirements
Operating in the defense sector requires navigating stringent security requirements that Powell acknowledges create unique challenges.
The company has achieved accreditation to handle classified information up to the Secret level through a partnership with Second Front Systems, a DevSecOps platform provider.
“We’re one of the few training platforms that is accredited to host classified information,” Powell noted. “That has taken a huge investment from our team to get there, but we think it is a huge advantage for us in the marketplace because we are able to serve some of the most critical missions that the Department of Defense is doing.”
This capability allows Ethos to work with sensitive operational data that traditional commercial learning platforms cannot access, Powell argued, giving the company a competitive advantage in defense applications.
AI and Human Oversight
Despite heavy reliance on artificial intelligence, Powell emphasized that human expertise remains central to the platform’s operation.
“Humans are a huge part of our platform and are always the final sign-off or the subject matter expert in relation to a new training program that we build,” he said.
The company uses AI primarily to reduce manual labor in course development, achieving what Powell claimed is “about a 90 percent reduction in the amount of time it takes, say, an instructional designer, to build a new course.” But human subject matter experts review all AI-generated content before deployment.
Powell also distinguished between different types of AI applications, noting that some automated functions are actually more reliable than human performance.
“If you think about some of the elements of the legal or contracting world where you really just need to follow a very strict process for how things are supposed to be done, AI systems are actually a good bit more effective in doing those things than humans because they don’t have the same error rate.”
The Broader Challenge
Powell’s background spans both education and business. After undergraduate work at the University of North Carolina and early experience with online learning platforms like Coursera, he helped build African Leadership University before attending Stanford Business School with his co-founder, Sasha Seymore.
His critique of traditional education extends beyond the military.
“Even the best universities in the world are not particularly good at the science of teaching and learning,” Powell argued, citing his Stanford experience. “Even a lot of those were not really leveraging technology in a highly effective, data-driven way.”
The military faces additional constraints, Powell noted, because “they have such stringent cybersecurity requirements that they can’t just go and pick up any software off the street and start using it for classified missions.”
This creates a situation where, as Air Force officials told him, “we haven’t changed the way that we train pilots in 40 years.”
Funding and Future Outlook
Ethos has raised $30 million across seed and Series A rounds, with backing from Norwest Venture Partners and Wesley Group, along with defense-focused angel investors.
Powell noted a significant shift in Silicon Valley’s attitude toward defense technology since the company’s early fundraising efforts in 2020.
“When we went out to raise our seed round in 2020, we did get kind of some confused investors who were just saying, why would you work in defense, or people would make ethical pushback,” Powell recalled. “I think that over the past few years, people have been persuaded by geopolitical events.”
He credited defense technology leaders like Palantir’s approach to “building software for the free world” with helping shift venture capital attitudes toward supporting military applications.
The Human Factor in Future Warfare
Despite his focus on AI-enabled training, Powell rejected the notion that technology will diminish the importance of human warfighters. Instead, he argued, individual operators will likely face greater responsibilities as they command multiple autonomous systems.
“What’s a Navy SEAL’s role going to be 10 years from now?” Powell asked. “They probably will have a whole set of autonomous combatants that are under their command. Now one person might be responsible for 100 weapon systems or a thousand.”
This expanding scope of responsibility, Powell contended, makes sophisticated training more critical, not less.
“Ensuring that that person is ready for every single possible situation they might find themselves in, when they’re facing a much more sophisticated adversary as well, I think it ends up meaning that the warfighter, the human warfighter, is substantially more powerful.”
The challenge, as Powell sees it, is ensuring military training can evolve as rapidly as the technology it supports. His bet is that AI-enabled platforms can finally make that possible, closing the gap between what warfighters need to know and what they actually know before it costs lives.
“I honestly couldn’t imagine a more impactful mission than ensuring that the millions of people in our armed forces are as prepared as possible to do their essential mission,” Powell said.
