a16zThe Person Who Runs HR For 2 Million Federal Workers
CHAPTERS
- 0:00 – 0:43
Making government “cool again”: why talent is the core constraint
Katherine Boyle frames the episode around revitalizing government capability, especially through better talent and technology adoption. Scott Kupor and Greg Barbaccia preview the central tension: rapid tech change vs. a federal workforce and systems not built to keep up.
- •Government modernization is increasingly viewed as a competitiveness issue
- •Technology is moving faster than government readiness, especially on AI
- •Talent—not just procurement—is positioned as the main bottleneck
- •Public service is pitched as high-impact work affecting hundreds of millions
- 0:43 – 1:40
Meet Scott Kupor (OPM) and Greg Barbaccia (US CIO): two private-sector-to-public leaders
Katherine introduces Scott Kupor as Director of OPM and Greg Barbaccia as the US CIO, emphasizing their atypical, tech-heavy backgrounds. The conversation sets up why this moment is drawing private-sector operators into public service.
- •Kupor comes from a16z leadership and investing experience
- •Barbaccia brings Palantir + national service background to federal tech leadership
- •A broader “renaissance” of tech talent entering government is highlighted
- •The episode will focus on practical levers: hiring, performance, and modernization
- 1:40 – 2:54
Why leave the private sector: fiscal pressure, tech acceleration, and a pro-change administration
Scott explains his motivation as a convergence of national fiscal realities and the government’s lack of readiness for fast-moving technology. Greg echoes that this administration is willing to disrupt existing ecosystems and push change across agencies.
- •Federal tech readiness gap is described as stark and urgent
- •Talent deficits are framed as a primary limiter on modernization
- •Leadership support (including AI-focused efforts) is cited as enabling reform
- •Greg emphasizes timing: a window where meaningful change is possible
- 2:54 – 5:12
What OPM actually does—and what “federal talent” means at 2+ million scale
Scott defines OPM as the federal government’s talent organization: hiring standards, performance management, and policies to attract and retain top people. He also outlines workforce scale and a projected reduction driven largely by voluntary programs.
- •OPM sets talent policies for the civilian federal workforce
- •Goals: attract, retain, and deploy high-quality talent for public outcomes
- •Workforce size context: ~2.4M civilian employees, trending toward ~2.1M
- •Reduction largely tied to voluntary/deferred resignation-type programs
- 5:12 – 6:18
What the US CIO does: tech policy, budgeting, and “one government” coordination
Greg clarifies that the federal CIO role differs from private-sector CIO roles, focusing on cross-agency policy and budgeting. A central priority is reducing siloed agency behavior and coordinating toward shared executive-branch goals.
- •CIO role: technology policy + budget decisions across the executive branch
- •“One government” effort aims to unify fragmented agency approaches
- •Cross-agency alignment requires galvanizing the CIO community
- •Silos are framed as a root cause of inefficiency and uneven execution
- 6:18 – 12:13
Culture shock: compliance complexity, lawyer-driven risk aversion, and endless oversight
Both leaders describe how government’s regulatory and oversight environment creates an extreme bias toward avoiding mistakes. Scott argues risk is treated as pass/fail rather than a spectrum with upside tradeoffs, reinforced by politics and oversight bodies.
- •Compliance/regulatory constraints make “move fast” approaches unrealistic
- •Risk is culturally over-weighted; upside is rarely part of decision calculus
- •Political incentives punish failures as moral/political indictments (e.g., “Solyndra” framing)
- •Oversight volume (GAO, OIG, Hill inquiries) discourages experimentation
- 12:13 – 17:31
Performance management: grade inflation, forced distributions, and building a merit culture
Scott presents updated performance-rating data showing near-universal “meets expectations or above,” arguing it breaks incentives for rewards and accountability. He describes introducing forced-distribution guidance for senior executives to begin shifting culture.
- •Performance ratings are highly inflated: ~99.7% receive 3+ (meets or better)
- •Overly positive ratings distort bonuses, promotions, and accountability
- •Forced-distribution guidance for SES aims to reintroduce differentiation
- •High-performance cultures attract high performers; merit needs to be rewarded
- 17:31 – 21:05
Recruiting and retention: mission as the differentiator, not lifetime employment
Greg and Scott explain why government can’t match Silicon Valley compensation but can win on mission and impact. They argue government has historically marketed the wrong value proposition (stability/tenure) instead of hard problems and national service.
- •Government competes on mission: scale of impact and problem difficulty
- •The “lifetime employment” pitch is called outdated and counterproductive
- •Top talent wants autonomy, modern tools, and meaningful outcomes
- •A key barrier: many tech roles devolve into managing contractors, not building
- 21:05 – 24:32
Targeting new talent pools: early-career hiring, technical managers, and public–private tours of duty
Scott outlines a strategy to prioritize early-career hiring to address a demographic imbalance and keep pace with AI-era skills. Both discuss bringing experienced private-sector technical managers into government (e.g., secondments) and making career movement fluid.
- •Federal workforce skew: only ~7% under 30; large share over 50
- •Early-career hiring is prioritized to build long-term capacity
- •Proposal: secondments for private-sector tech leaders to manage and mentor teams
- •Rejecting a false dichotomy: careers can alternate between public and private sectors
- 24:32 – 28:48
Fixing the “non-technical gatekeeper” problem: better screening, contracting, and skills assessment
Greg explains how lack of technical expertise in hiring and procurement leads to bloated, mis-scoped tech contracts. Scott adds that resume screening often relies on self-attestation and describes moving back toward functional assessments now that old constraints have been lifted.
- •Non-technical screeners struggle to evaluate engineers and technical roles
- •Procurement suffers similarly: approvers can’t sanity-check tech scope/cost like physical goods
- •Legacy contracting dynamics enable large, poorly structured long-term tech deals
- •Shift toward functional assessments (e.g., coding tests) to validate real capability
- 28:48 – 36:29
AI and modernization in practice: shifting from process worship to outcome-driven automation
Katherine pushes on how to reorient government around AI similarly to how companies reoriented after ChatGPT. Greg argues government over-values ritual and box-checking; Scott emphasizes micro-adoption—getting tools into hands, teaching safe use, and shipping small wins rather than writing long plans.
- •Government often rewards completing process, not delivering outcomes
- •AI is framed as the next modernization wave (like email and spreadsheets were)
- •Tactical approach: deploy tools broadly, encourage small efficiency gains, iterate
- •Training and risk messaging should be lightweight and adoption-oriented (avoid memo bureaucracy)
- 36:29 – 40:38
Defining success: operational efficiency, better citizen experience, and integrated data across agencies
Both leaders describe how they’ll measure impact during their tours of duty. Scott wants operational efficiency to become a first-class metric; Greg targets a unified citizen experience and responsible data sharing that breaks down agency silos.
- •Scott: reward high-quality services delivered at lower cost (stewardship of taxpayer dollars)
- •Scott: make government attractive to early-career talent and valued by private employers afterward
- •Greg: simplify the fragmented “web of government websites” into citizen-centric experiences
- •Greg: consent-based data sharing and cross-agency intelligence from integrated datasets
- 40:38 – 42:56
Lightning round: myths to retire, high-leverage tweaks, key metrics, and winning the future
In rapid-fire format, they challenge common assumptions about government tech capability and workplace tools. They propose small but high-impact shifts (risk thinking and modernizing legacy compliance), share metrics and reading recommendations, and close with their “win the future” completion.
- •Myths: government can’t lead in tech; government tools are necessarily terrible
- •Process tweak: treat risk as a spectrum; update pre-computer compliance assumptions
- •Metrics: operational efficiency; rework percentage (doing the same work repeatedly)
- •Close: “Win the AI race” and “Remain undistracted”
