Winning the AI Race Part 1: Michael Kratsios, Kelly Loeffler, Shyam Sankar, Chris Power

Strategic Vision and Administrative Priorities

• The Trump administration has designated winning the global AI race against China as a primary national security and economic priority, aiming for US dominance in all industries.
• President Trump signed an executive order in his first week directing the creation of an AI action plan involving the Office of Science and Technology Policy (OSTP) and the National Security Advisor.
• The administration views the current trajectory of AI as an "infinite game" where the US must harness technological superiority to drive economic vibrancy and military supremacy.
• A central tenet of the administration is rejecting the narrative that AI destroys jobs, instead framing it as a massive engine for job creation and industrial re-industrialization.
• The administration's approach involves a "pivot back" to optimism, celebrating American technology and people while contrasting this with historical global competitors who have created existential risks through mistrust.

The AI Action Plan: Three Pillars

• Innovation: The plan focuses on out-innovating global competitors through deregulation, avoiding the creation of rigid laws that could hinder the integration of AI into other technologies (e.g., drones, medical diagnostics).
• Infrastructure: The administration prioritizes accelerating the construction of AI-specific infrastructure, including data centers, energy generation, and chip manufacturing, within the United States.
• Ecosystem: The goal is to build the largest possible AI ecosystem by dominating the "AI stack" for the world, mirroring the success of Silicon Valley platforms with the most developers and applications.
• Concrete Actions: The executive branch has released a plan containing 90 concrete actions designed to be accomplished within the next six to nine months to accelerate US leadership.
• Public Feedback: The drafting process solicited over 10,000 responses from diverse sectors, including Hollywood and technology companies, to ensure the plan addressed broad national needs.

Data, Regulation, and Energy

• The government plans to leverage its vast datasets (e.g., Department of Energy national labs) to drive scientific discovery in material science and medicine, moving beyond simple LLM coding applications.
• The administration is exploring the use of categorical exclusions for AI-related activities on federal lands to accelerate the deployment of data centers and power generation.
• A $150 million initiative under the Bipartisan Budget Bill is directed at the Department of Energy to build an "AI for Science" program to clean, homogenize, and unlock government data for private industry.
• Officials acknowledge a threat to national security from a potential "patchwork" of over 1,000 proposed or passed state-level AI regulations, though the current action plan focuses on executive branch actions rather than federal preemption.
• The administration emphasizes that "unlocking energy" is a prerequisite for winning the AI race, noting that China is on track to triple its energy generation capacity compared to the US by 2030.
• To support re-industrialization, the administration argues for leveling the playing field regarding the cost of energy and raw materials, which are heavily subsidized in China but costly in the US.

Workforce Development and Reshoring

• The strategy prioritizes upskilling the existing American workforce and attracting the "greatest scientists and engineers" globally, while also reskilling non-traditional workers for manufacturing roles.
• Hadrian CEO Chris Power highlights a critical shortage of skilled trades (e.g., welders, machinists), noting that US defense manufacturing relies on a retiring workforce of older skilled Americans.
• Hadrian reports a 4x jump in manufacturing productivity and a 10x jump in workforce productivity by using AI-powered factories to automate complex tasks previously requiring decades of experience.
• The "Tech Fellows Program" at Palantir aims to identify and credential "autodidact" blue-collar workers with mechanical intuition, training them in four weeks to lead AI implementation in their organizations.
• Small business owners (SBA) report that AI is enabling the return of industries previously offshored, such as bicycle manufacturing in Seymour, Indiana, creating 60 high-paying jobs in a single factory.
• The SBA has rewritten rules to allow small businesses to use loan proceeds for AI implementation and advanced manufacturing, reversing previous bans under the Biden administration.

Industry-Specific Applications and Case Studies

• Manufacturing: Hadrian is building "full-stack AI-powered factories" in the US, with Factory 2 in LA and a new gigafactory in Arizona launching in six months, aiming to create 350+ new AI-powered jobs.
• Healthcare: AI tools at Tampa General Hospital and UP Health System have been shown to halve the time nurses spend on paperwork, allowing more time for bedside patient care and earlier disease detection.
• Defense: PRL Industries utilizes AI to reduce quoting processes from three days to minutes, freeing engineers to solve production problems rather than manage administrative data.
• Energy: Gecko Robotics uses robots to gather "physical data" (e.g., from power plants and mining sites) to train AI models that predict asset failure, extending the useful life of grid infrastructure by up to 30 years.
• Battery Production: Panasonic Energy uses AI to analyze historical maintenance data to predict equipment breakdowns, shifting from reactive to predictive maintenance and retaining technician confidence.

The Role of Small Business and Startups

• The SBA is on pace for a record year with 2,000 Main Street loans issued weekly, with 60% of $21 billion lent this year going to companies with one to five employees.
• The administration aims to double the size of SBA loans for purchasing advanced technology and equipment, focusing on critical industries like metals, minerals, and medical devices.
• SBA Administrator Kelly Loeffler notes that underwriting standards have been restored to pre-pandemic levels, resulting in a loss ratio of 3% or less and eliminating the need for taxpayer subsidies.
• Y Combinator's Paul Bluhite predicts a 10x to 100x increase in startups as "vibe coding" and natural language interfaces lower the barrier to entry, shifting coding from a specialized skill to a common language.
• Palantir's Sean Toffey argues that AI is not displacing workers but giving "superpowers" to the average American, enabling small businesses to compete with large incumbents by leveling the playing field on marketing, legal, and operational expertise.

Future Outlook and Investment

• Paul Bluhite anticipates that while foundation models will require massive capital (limiting the number of providers), open source will remain a critical option to prevent monopolies and ensure alignment.
• The conversation shifts toward "physical AI" and robotics, with a notable increase in robot arm deployments at recent demo days as technology curves make complex automation tractable.
• The administration and industry leaders agree that the next wave of wealth creation will come from activating historically intractable problems (e.g., drug discovery, complex manufacturing) rather than just displacing old workflows.
• The SBA is exploring equity structures or warrants for investments in critical defense technologies to provide taxpayers with upside potential while managing risk, moving away from "plain vanilla" loan structures.
• A key trend identified is the "new collar boom," where traditional four-year degrees are becoming less relevant than practical, hands-on technical skills in the era of AI-augmented manufacturing.