AI Isn’t the Problem—Big Tech Is (w /Josh Tyrangiel) | How to Fix It
YouTube · 5g3dv46M5Xo
Quick Read
Summary
Takeaways
- ❖AI's true potential for public good is often overshadowed by Big Tech's focus on ad-driven, customized content.
- ❖Operation Warp Speed demonstrated AI's power to manage complex logistics, rapidly distributing COVID-19 vaccines by untangling vast data sets.
- ❖Government's inability to effectively procure and deploy software, due to rigid rules and political scrutiny, is a major impediment to AI adoption.
- ❖Successful AI implementation in government often requires a 'nothing to see here' approach, bypassing bureaucratic resistance through quiet, incremental changes.
- ❖While AI can boost productivity and make some jobs obsolete, a gradual 15-20 year transition could allow for labor market adjustment and new job creation.
Insights
1Operation Warp Speed: A Blueprint for AI in Government
General Gus Perna successfully leveraged Palantir's AI-powered ontology building to create an 'end-to-end god view' of the COVID-19 vaccine supply chain. This allowed for rapid data integration, cleaning, and real-time monitoring of vaccine production, distribution, and storage, leading to an unprecedentedly fast rollout that bypassed typical bureaucratic delays.
Palantir connected pharmaceutical companies to the smallest rural pharmacies, managing data from hundreds of inputs to ensure vials, bags, and vaccines reached arms. This was achieved in just six to eight weeks.
2Government's Software Deployment Crisis (Gaul's Law)
The U.S. government's procurement system is designed for tangible hardware (like tanks) rather than evolving software. This leads to slow, expensive, and often ineffective software development because contracts are rigid, political accountability punishes adaptation, and lawmakers lack tech fluency. Eric Schmidt found this deeply frustrating.
Congress members are among the least technologically educated, and the system prevents changes to contracts even when user needs evolve, leading to 'slow and wrong' software. Jennifer Pahlka described a 'talent crisis' due to these restrictive rules.
3Local AI Success: East Lansing's Recycling Program
East Lansing implemented Prairie Robotics' AI system to improve recycling rates. Cameras on trucks identified contaminants, and residents received personalized postcards (some with photos of their actual contaminants) guiding them on proper recycling. This 'nothing to see here' approach, focusing on effective policy rather than hyping AI, led to massive gains and high public approval.
The $8,000 system used AI to identify recyclable items and contaminants from hundreds of thousands of photos. Postcards with low-res photos of actual contaminants led to significant improvements and 85% public approval.
4IRS Modernization: Quiet AI Integration
The IRS is quietly using AI to modernize its operations, including creating a ChatGPT-like search engine for customer service agents and migrating its ancient Individual Master File (IMF) to modern coding languages. This is done within strict legal confines, ensuring human oversight for 'inherently governmental' functions like tax record evaluation, avoiding political backlash.
IRS Commissioner Danny Werfel and CTO Kashif Panda are using AI to make customer service manuals searchable and update the 1960s-era IMF, which holds all taxpayer records. They emphasize 'human in the loop' to maintain legal compliance and public trust.
5The Perils of Politicized AI Initiatives (Doge)
Politically driven initiatives, like the Trump administration's 'Doge' project, often fail because they prioritize cuts and political points over understanding departmental needs or citizen services. These efforts lack respect for existing rules and regulations, leading to destructive outcomes rather than genuine improvement.
Doge hires had AI qualifications but 'didn't bother to learn anything about the departments they were trying to nominally help' or the services provided to taxpayers/veterans. The focus was on 'cuts' rather than thoughtful AI integration.
Bottom Line
The 'nothing to see here' strategy for AI adoption in government can be highly effective.
By quietly implementing AI for tangible service improvements without hyping the technology, local and federal agencies can bypass political scrutiny and bureaucratic resistance, achieving significant gains before opposition mobilizes.
Government leaders should prioritize incremental, user-focused AI deployments and communicate their benefits in terms of service improvement rather than technological revolution, fostering acceptance and demonstrating value.
The biggest barrier to government AI is not the technology, but the government's outdated procurement and regulatory framework.
Current rules, designed for hardware, stifle the iterative, evolving nature of software and AI. This prevents agile development, makes innovation costly, and discourages talented tech professionals from joining public service.
Policymakers should advocate for procurement reform that allows for flexible, performance-based contracts for software, encourages rapid prototyping, and prioritizes user feedback, enabling government to leverage modern tech effectively.
Key Concepts
Gaul's Law
Every complicated system that works started from a simple system that works. This law highlights why government struggles with software: it attempts to build complex software systems from scratch without understanding the iterative, evolving nature of software development, unlike hardware procurement.
Human in the Loop
This concept emphasizes the necessity of human oversight and guidance in AI systems. It ensures that AI serves as a tool to augment human capabilities and decision-making, rather than replacing critical human judgment, especially in sensitive areas like government services or tax records.
Lessons
- Government leaders must develop a foundational understanding of AI's capabilities and limitations to speak credibly about its deployment and avoid politically motivated pitfalls.
- Prioritize AI implementation in areas that demonstrably improve citizen services, such as Veterans Affairs or tax filing, to build public trust and demonstrate tangible benefits.
- Advocate for procurement reform that supports agile software development, allowing for iterative improvements and responsiveness to user needs, rather than rigid, 'one price, one outcome' contracts.
Notable Moments
Eric Schmidt, after four years on the Defense Innovation Board, became 'despondent' over the government's inability to deploy effective software.
This highlights the profound systemic challenges within government bureaucracy that even highly influential tech leaders struggle to overcome, underscoring the depth of the problem.
Engineers working on Operation Warp Speed felt the 'extraordinary things going away' as bureaucracy returned once the immediate crisis subsided.
This illustrates how crisis conditions can temporarily suspend bureaucratic hurdles, allowing for rapid innovation, but without fundamental reform, inertia quickly reasserts itself, stifling ongoing progress.
Quotes
"The thesis is that basically that the tech needs to be separated from the tech companies because if you believe the tech companies and what they want you to do with AI, they're going to spoon feed you AI that serves you better ads, gives you more customized content of all kinds, safe and not safe for work. And basically it's not that different to what happened with social media where you're going to be dazzled at first and then suddenly you're going to realize you work for them."
"Every complicated system that works started from a simple system that works."
"You will never have effective deployment of AI in government until the government starts to understand the best practices for deploying software."
"In the year of our Lord, 2026, you and I just filed taxes. We have to guess what we owe the government under penalty of prosecution. This is insane."
"One thing Democrats have to learn is like defending obsolete jobs isn't really helping anybody."
Q&A
Recent Questions
Related Episodes
The GPT Moment for Robotics Is Here
"Physical Intelligence is pioneering general-purpose robotics, leveraging cloud-hosted AI models and cross-embodiment data to enable a 'Cambrian explosion' of vertical robotics companies."
I put 80% of my money in the S&P after a billionaire investor told me not to
"This episode explores how genetics influence investing behavior, the importance of self-awareness in career and investment choices, and groundbreaking business ideas in AI, defense tech, and personalized medicine."
The Joe Budden Podcast Episode 926 | Let’s See That Snapper
"The hosts dissect the collapse of Spirit Airlines, the predatory nature of music contracts, and the weaponization of private sexual preferences in public disputes, all while celebrating the Knicks' playoff dominance."
Hakeem Jeffries Humiliated as Ron DeSantis’ Hakeem Impression Goes Viral
"Dave Rubin critiques Democratic policies and media narratives, contrasting the perceived decline of progressive-led American cities with the flourishing of conservative states and El Salvador, while highlighting Ron DeSantis's viral impression of Hakeem Jeffries."