Big Technology Podcast
Episode: Booz Allen CTO: Can AI Fix The Government — With Bill Vass
Host: Alex Kantrowitz
Guest: Bill Vass, CTO of Booz Allen, former Amazon executive
Date: September 10, 2025
Episode Overview
In this wide-ranging episode, Alex Kantrowitz sits down with Bill Vass, CTO of Booz Allen and former Amazon executive, to explore how AI is transforming the efficiency, effectiveness, and future of government technology. The conversation covers everything from government IT redundancies and contract models to real-world AI deployments, quantum computing, robotics, and lessons from the tech industry's rapid evolution. Vass brings an insider’s perspective from decades straddling public and private sector technology leadership.
Key Discussion Points and Insights
1. Booz Allen’s Role in Government Tech (01:14–01:56)
- Booz Allen is now primarily a government technology contractor with 22,000 engineers, including thousands specializing in AI and cybersecurity.
- Booz Allen builds everything from quantum hardware and intelligence satellites to exploratory projects like 3D-printed organs.
- "Just a bunch of software developers that do everything from building hardware qubits for the government to running the GPS satellites... So it's a pretty broad range of tech, pretty exciting actually."
— Bill Vass [01:34]
2. Government IT Redundancy and Modernization (01:56–03:57)
- Legacy IT and agency duplication: Vass recalls the Pentagon running 255 different systems in the 1990s due to "stove pipe" organizations and political barriers to consolidation.
- Consolidation efforts are ongoing but hampered by agency independence and political interests.
- "Some of it is politics... different divisions that want to do it themselves... I think there's just a lot of places like that where that kind of stuff shouldn't be tolerated."
— Bill Vass [03:28]
- "Some of it is politics... different divisions that want to do it themselves... I think there's just a lot of places like that where that kind of stuff shouldn't be tolerated."
3. Contract Reform and Outcome-Based Models (05:39–08:00)
- Shift from time and materials to outcome-based contracts: The government is moving from contracts that pay for hours worked to those paying for defined results—a change Vass praises as more efficient and taxpayer-friendly.
- "Migrating from on-prem to the cloud should be an outcome-based contract...When you know how to do something, outcome based makes a lot of sense."
— Bill Vass [06:39]
- "Migrating from on-prem to the cloud should be an outcome-based contract...When you know how to do something, outcome based makes a lot of sense."
4. Stereotypes vs. Reality: State of Government Technology (12:28–18:27)
- Public perception: Many believe government tech is outdated, with processes stuck on systems like Windows 95.
- Reality is a mixed bag: Advanced examples include GPS, Mars rover, intelligence satellites; legacy exists due to funding choices and politics, not a lack of technical capability.
- "I see more cutting edge technology in the government often than I see in Silicon Valley and...I also see things where the government's partnered up with Silicon Valley to deliver things."
— Bill Vass [16:56] - "Whenever you see a bad technology decision, it's always politics."
— Bill Vass [20:42]
- "I see more cutting edge technology in the government often than I see in Silicon Valley and...I also see things where the government's partnered up with Silicon Valley to deliver things."
5. AI in Government Practice (21:16–26:11)
- Real Deployments:
- On the International Space Station: LLaMA runs locally, providing astronauts with zero-latency AI assistance for troubleshooting and referencing manuals.
- Veterans Administration: Large language models (LLMs) accelerate claims processing from hours to seconds.
- Autonomous Systems: Debate between rules-based and LLM-based autonomy for vehicles, drones, and ISR systems.
- AI for code development: Booz Allen and government teams use Copilot, Claude, and similar tools.
- Limits & guardrails: Critical systems require human oversight; model outputs can vary, and costs are still high.
- "All ML is just math... vectors and tensors... It's not magic, it's just math."
— Bill Vass [25:15]
- "All ML is just math... vectors and tensors... It's not magic, it's just math."
6. The Hallucination and Overreliance Debate (26:11–31:11)
- Skepticism about LLM trustworthiness: Critics worry about hallucination and cognitive “atrophy.”
- Vass’s take: LLMs are tools like calculators or knives—helpful, imperfect, and to be paired with human judgment.
- "You can cut yourself with a knife in the kitchen... We've lost the talent of tearing things apart with our hands because we've invented knives. I think that's overblown... They need to understand those limitations."
— Bill Vass [29:22]
- "You can cut yourself with a knife in the kitchen... We've lost the talent of tearing things apart with our hands because we've invented knives. I think that's overblown... They need to understand those limitations."
7. Dreaming Big: AI’s Promise for Government Services (31:11–33:09)
- Future vision: AI could massively improve citizen services, efficiency, and speed—though costs (mostly for GPUs) and ROI questions remain.
- Human touch: Tech can’t fully replace people, especially for complex or sensitive government services.
- "I would imagine a world that’s got better citizen services that can deliver things faster... But at some point in time a citizen should expect to talk to a person."
— Bill Vass [32:02, 33:08]
- "I would imagine a world that’s got better citizen services that can deliver things faster... But at some point in time a citizen should expect to talk to a person."
8. Global Stakes: Watching Saudi Arabia’s Sovereign AI Push (33:09–35:35)
- Saudi Arabia is emerging as a global AI experiment with huge investments from Nvidia, Amazon, and others.
- Will the world watch?
- "There's a tremendous amount of brain trust happening in Saudi Arabia and investment there...I think that's a good investment and the right thing to do to transform that region."
— Bill Vass [34:33]
- "There's a tremendous amount of brain trust happening in Saudi Arabia and investment there...I think that's a good investment and the right thing to do to transform that region."
9. Importing Leadership from Tech to Government (36:39–38:33)
- Moving Amazon culture to the US government: Notions like “bias for action,” "think big," and "dive deep" are becoming more common in the public sector, but “customer obsession” for citizens is still lagging.
- "The government’s got a lot more bias for action right now... People are willing to think big... but I don't think the government is as customer obsessed as it should be."
— Bill Vass [37:38]
- "The government’s got a lot more bias for action right now... People are willing to think big... but I don't think the government is as customer obsessed as it should be."
10. The Reality and Future of Autonomous Driving (39:59–44:49)
- Current state: Vass owns Teslas and tests Full Self Driving (FSD), but describes it as "entertaining" rather than 100% reliable.
- Challenges and edge cases: Training requires vast synthetic simulation (Omniverse, Unity, Unreal). Hard for models to interpret subtle cues, like the direction of a car's wheels at a stop sign or informal hand signals.
- Prediction: "You’ll start to see real autonomous driving over the next five years... There’s still a lot of complexity."
— Bill Vass [41:55]
11. Robotics Race: US vs. China (45:10–49:18)
- US is competitive, not behind—innovation hinges on public-private partnerships. The government now sees AI as a “pacing threat” issue with China and is refocusing R&D.
- "I was worried about us falling behind the Chinese and the government... so Booz Allen... was a great way, I felt, to more directly influence and improve that technology."
— Bill Vass [47:15]
12. Quantum Computing: Hype and Reality (49:18–59:37)
- Where we are: Quantum machines work but are too noisy—error correction is the bottleneck.
- “The biggest thing that it will enable first... is going to be material sciences and chemistry first... For example, you can simulate ammonia’s Hamiltonian in about three minutes on a quantum computer versus the age of the universe on all the world’s classical compute.”
— Bill Vass [54:05]
- “The biggest thing that it will enable first... is going to be material sciences and chemistry first... For example, you can simulate ammonia’s Hamiltonian in about three minutes on a quantum computer versus the age of the universe on all the world’s classical compute.”
- Timelines:
- Football-field scale machines, first applications in ~2032.
- Urgency for US government and companies to shift to quantum-safe cryptography as quantum will be able to break RSA-class encryption by ~2040.
- Competitive implications:
- "The country that has this first will have a tremendous lead... in material sciences at first, but later in cryptographic sciences."
— Bill Vass [58:23]
- "The country that has this first will have a tremendous lead... in material sciences at first, but later in cryptographic sciences."
13. Lessons from Sun Microsystems and Tech Industry Survival (59:54–65:07)
- Sun’s legacy: Invented vital tech but struggled with transitions, open source, and making products accessible to non-engineers.
- "The only constant is change in this industry... You must be your own best cannibal... If you don’t replace [your own tech], your competition will."
— Bill Vass [60:52, 61:07]
- "The only constant is change in this industry... You must be your own best cannibal... If you don’t replace [your own tech], your competition will."
- Advice: Don’t let “the best be the enemy of the better,” move fast, and stay customer-focused, not just engineer-focused.
Notable Quotes & Memorable Moments
- On Government Tech Perception:
- "I think categorizing it as all government technology is bad is absolutely wrong. A lot of it is quite good. A lot of it is quite impressive."
— Bill Vass [15:07]
- "I think categorizing it as all government technology is bad is absolutely wrong. A lot of it is quite good. A lot of it is quite impressive."
- On AI Hallucination:
- "I must have spent 20 minutes on Google looking for [two projectors] and realized it had made them up...That's really though, the important thing to understand on how these tools work."
— Bill Vass [29:52]
- "I must have spent 20 minutes on Google looking for [two projectors] and realized it had made them up...That's really though, the important thing to understand on how these tools work."
- On Quantum’s Potential:
- "If you took all the iPhones and all the laptops and all the Android phones and all the cloud computers on Earth...it would run for longer than the history of the universe...With a thousand error rector qubits, it would take about three minutes."
— Bill Vass [54:24]
- "If you took all the iPhones and all the laptops and all the Android phones and all the cloud computers on Earth...it would run for longer than the history of the universe...With a thousand error rector qubits, it would take about three minutes."
Important Segment Timestamps
- Booz Allen’s Present and Scope: 01:14–01:56
- Redundancy & Pentagon Example: 01:56–03:57
- Contracting Models: 05:39–08:00
- Perception vs. Reality in Gov IT: 12:28–18:27
- Technology Decision and Politics: 19:37–20:42
- Government AI Deployments: 21:16–26:11
- AI Hallucination and Reliability: 26:11–31:11
- Dream Scenario for AI in Public Sector: 31:11–33:09
- Saudi Arabia’s AI Strategy: 33:09–35:35
- Bringing Tech Culture to Government: 36:39–38:33
- Autonomous Driving and Simulation: 39:59–44:49
- China-US Robotics Race: 45:10–49:18
- Quantum: Reality, Timeline, and Use Cases: 49:18–59:37
- Lessons from Sun Microsystems: 59:54–65:07
Concluding Tone
Throughout the episode, the conversation is open and realistic but hopeful. Vass’s tone is pragmatic—acknowledging frustrations with bureaucracy and inertia but also regularly emphasizing the massive technical talent and innovation that exists inside the US government. There’s energy, optimism about AI’s potential, and a call to move fast, whether in government or industry.
This episode is a comprehensive resource for anyone interested in the intersection of AI, public sector modernization, real-world tech deployment, and the global race for technological leadership.
