By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
a16zThe Software Crisis Behind America's Infrastructure
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
CHAPTERS
- 0:00 – 3:23
Phillip's Path from Germany to Silicon Valley
Leila Hay asks Phillip Buckendorf about the moment he left Germany. He describes growing up there, being frustrated by its approach to energy, consulting culture and obstacles to building companies, and moving to a Palo Alto hacker house.
- •Buckendorf cites Germany's nuclear exit and reliance on Russian gas as part of his frustration
- •He found friends glorifying consulting while government made starting companies hard
- •He went to Palo Alto, lived in a hacker house and never moved back
- •He credits the American ecosystem, ethos and energy for what he built
- 3:23 – 7:09
Founding ASI: Outsiders Discover Ancient Software
Buckendorf explains that he and his co-founders came from autonomous driving and asked which other transportation modes needed better software. Visiting operation centers, they found extremely outdated software, which set ASI's mission. Leonard Kosinski then describes his 30-year military career and why ASI drew him.
- •The founders were outsiders: not pilots, defense insiders or logisticians
- •They avoided crowded autonomous driving and toured maritime and air operation centers
- •They expected science fiction but found the most ancient software possible
- •Kosinski, a retired three-star general and former Joint Staff logistics director, was drawn by ASI's focus on data
- •He saw the Department of Defense struggle to access and optimize data
- 7:09 – 11:28
Aviation's Interlinked Crisis: Staffing, Software, Infrastructure
Asked to paint a picture of aviation, Buckendorf says flying remains the safest mode but faces staffing shortages, faltering legacy software and outdated infrastructure. He argues these are interconnected and fundamentally software problems, and notes new trainees expect modern tools.
- •Staffing shortage stems from retirements, slowed training during COVID and difficulty attracting top talent
- •Legacy software failures cause massive, consequential outages
- •Better software speeds training, raises productivity and could allow higher pay
- •A new generation raised on iPads and Google Maps is trained on systems built for IBM green screens
- 11:28 – 15:28
The Philosophy Problem in Software Modernization
Prompted by the recent DC collision, Buckendorf frames legacy technology as a philosophy problem with three issues: software coupled to compute, hardware-style development with huge documentation and decade-long builds, and inability to attract top engineers. Kosinski says the Department of Defense faces the same ecosystem problem.
- •Software tightly coupled to compute makes updates facility-by-facility and risky
- •Programs spend tens of millions on documentation, then a decade building from scratch, and deliver outdated systems
- •Top software engineers don't want to work from lists of thousands of requirements
- •Acquisition frameworks incentivize these flawed philosophies
- •Kosinski says DoD logistics systems were not designed to connect and need clean-sheet, iterative software
- 15:28 – 20:16
Funding Modernization and Buying Proven Software
On the urgency of fixing the status quo, Buckendorf says the administration has set the mandate and Congress should fund FAA modernization while ensuring past mistakes aren't repeated. He argues for buying commercially proven software, citing airlines and the Air Force as examples.
- •Buckendorf cites strong bipartisan support for modernizing air traffic control
- •He says funding must come with guidelines to avoid repeating 10-year custom builds
- •Commercially deployed software is faster and safer because it is already proven
- •ASI deployed with the Air Force within months because its software already worked commercially
- •Kosinski adds that the military daily relies on commercial trucking, rail, air and sea
- 20:16 – 24:04
What Contested Logistics Means
Kosinski defines contested logistics using the analogy of ordering online: logistics is unnoticed until disrupted, and adversaries target supply chain vulnerabilities. He argues logistics is itself a weapon system and a competitive advantage, which Hay compares to electricity or water.
- •Contested logistics covers weather, maintenance and adversary action, requiring prediction and optimization
- •Moving forces requires a whole supply 'tail' of food, munitions and more
- •Kosinski calls logistics a weapon system that is either an advantage or a greatest weakness
- •Examples include pandemic shortages and support to Ukraine
- •Buckendorf says 'silent software' gets neglected versus visible weapon systems
- 24:04 – 25:54
Why Logistics Is Underserved: Human Nature and Just-in-Time
Hay asks why logistics gets so little attention despite the pandemic. Kosinski points to human nature, decades without global war, and a just-in-time mindset that treats logistics as a cost rather than a competitive advantage.
- •People ease back to the status quo after crises like the pandemic
- •Decades of comfort meant no global war pressuring supply chains
- •Just-in-time fails if the part can't be obtained
- •Kosinski says logistics should be seen as a source of profit or deterrence, not a cost
- 25:54 – 30:15
Dual-Use Logistics Software and Allied Collaboration
Buckendorf explains why logistics is the flagship dual-use domain: much defense capacity sits in the private sector and shares ports and infrastructure, and adversaries push their own port software. Kosinski adds that NATO is moving from collective defense toward collective logistics.
- •Some capabilities like missiles are sector-specific, but logistics benefits from shared software
- •Buckendorf says Chinese software is being given away free in allied ports
- •Shared software stacks let sectors share data before a crisis rather than during one
- •ASI aims to enable collaboration between private and government sectors
- •Kosinski says NATO is shifting to collective logistics because no nation can do it alone
- 30:15 – 34:21
Modernizing DoD Logistics: From Dashboards to Prediction
Asked what must change for DoD logistics, Buckendorf says recent modernization mostly added dashboards, and software must now predict what is about to happen. He outlines evolution from workstations to connected data fusion to prediction machines, plus the need for collaboration and proactive funding.
- •Buckendorf says anticipation is the new high ground in an uncertain world
- •Software evolved from standalone compute to internet and IoT data fusion to prediction machines
- •Existing capacity across the US, allies and private sector must be coordinated rather than only building more ships
- •Sectors should be funded proactively, not after things fall apart
- 34:21 – 36:07
AI Decision Tools and Training Joint Logisticians
Kosinski stresses funding and prioritization, then describes the challenge of creating joint logisticians who understand air, sea and ground logistics, a process taking decades. He argues AI decision tools can accelerate this by suggesting multimodal options.
- •Logistics is fundamental and needs policy-level funding and prioritization
- •Joint logisticians must master multiple service perspectives over many years
- •AI decision tools can offer multimodal options such as ship or air
- •Training should continue, but technology shouldn't depend on one all-knowing individual
- 36:07 – 39:00
The Future of AI in Logistics: Ten Years Out
Asked what problems get solved in ten years if this is done right, Buckendorf says more can be done with existing capacity and networks can be hardened against disruption through rerouting and anticipation. Kosinski adds that strong logistics provides the deterrence needed to prevent war.
- •More output from existing capacity
- •Hardened logistics networks with fast rerouting amid geopolitical, sanction and climate uncertainty
- •Predictive capability protects both warfighters and civilians
- •Kosinski says sustainment is fundamental to deterrence and adversaries know it
- 39:00 – 43:37
Risks of Not Modernizing: Adversaries, Vulnerable Nodes and Urgency
Asked what is at stake, Kosinski cites urgency, "The 100-Year Marathon," and the loss of uncontested movement even at home, noting DoD logistics runs on a 40-year-old system. Buckendorf points to geography and adversary strategies, and both cite Colonial Pipeline, the Suez Canal and threats to infrastructure nodes.
- •Kosinski says adversaries like China have studied US supply chain vulnerabilities for a long time
- •Ports, rail and supply chains are at risk from hypersonic, cyber and space threats
- •Buckendorf says oceans protect the homeland but power projection depends on logistics, which adversaries aim to cripple
- •Colonial Pipeline and Suez Canal show real vulnerabilities
- •Predictive, AI-enabled logistics and threat detection at every node are needed
