a16zMark Cuban on the NBA, Cost Plus Drugs, and How to Fix Politics
CHAPTERS
Competing in ideas: group chats, dissent, and independent politics
Cuban describes why he spends time in group chats: to learn, challenge assumptions, and seek real reasoning instead of performative social-media outrage. He frames himself as politically independent, focused on issue-by-issue thinking and understanding the “trust” layer beneath people’s beliefs.
- •Group chats as a high-signal place for intellectual disagreement vs. “you suck” discourse
- •Being the dissenting voice to uncover underlying assumptions
- •Political identity: independent, not party-loyal; has voted for Republicans
- •Many political disagreements reduce to trust in institutions and information sources
Platforms, algorithms, and the “salesperson” advantage in modern politics
Cuban connects business instincts—PR, sales, and distribution—to political success, using Trump as an example of algorithm-aware messaging and “flooding the zone.” He argues Democrats struggle because they lack strong sellers and fail to package messages for how people actually consume media.
- •Social media created a direct distribution advantage for public business figures
- •Trump as a master of PR/sales and algorithmic reach (“flood the zone”)
- •Democrats’ messaging weakness: fewer people who know how to sell
- •AI will accelerate content volume and targeting, changing political/media playbooks
AI-generated media and the near-term future of attention
The conversation turns to how AI will transform content creation and consumption, including rapid video generation and potential “agentic” content flooding. Cuban predicts a pivotal few years as audiences decide whether to disengage due to synthetic content or embrace new formats and norms.
- •AI enables near-instant generation and mass personalization of content
- •Risk: social platforms become less “social” as synthetic media overwhelms feeds
- •Micro-events and comments can shift sentiment instantly; polling becomes less useful
- •Creators will iterate faster, producing new kinds of media experiences
What comes after today’s social platforms: fragmentation, new media, and less rage bait
Cuban expects continued fragmentation (e.g., BlueSky spikes) but suggests the next wave may look more like “media” than “social.” He’s optimistic that LLMs as fact-checking companions (e.g., Grok/ChatGPT/Perplexity) could reduce misinformation and, therefore, rage bait—creating room for higher-quality content models.
- •Platform fragmentation creates silos; “media” may replace “social” as the core concept
- •Creators will amplify output, but distribution control remains the bottleneck
- •LLMs can counter expected-answer bias and reduce misinformation
- •Less misinformation could reduce rage bait and change how audiences connect to content
SEC experience and why privacy products still matter
Cuban recounts being charged by the SEC and how statements were taken out of context, motivating him to build a truly private communication tool (originally “Dust”). The product persists today primarily as an internal company communications solution.
- •SEC case as a catalyst for building privacy-first communications
- •Desire for “truly private” messaging to avoid context manipulation
- •The product still exists, now used mostly for internal employee comms
- •Privacy remains a strategic need even as mainstream platforms dominate
Fixing Democratic messaging: sell “a better life,” not abstract threats
Cuban argues Democrats over-index on projecting worst-case scenarios and trigger-word politics rather than addressing immediate cost-of-living realities. He shares an anecdote—saying “I don’t watch the news” in Indiana—that revealed deep public fatigue with media and apocalyptic framing.
- •Democrats should lead with tangible daily issues (prices, housing, education choices)
- •Pep-rally rhetoric energizes base but doesn’t persuade the median voter
- •Public exhaustion with news/media narratives is a political opportunity
- •Political success requires clear goals and messaging tied to lived experience
Reducing inequality via employee ownership and “appreciable assets”
Instead of “tax the billionaires” as an end in itself, Cuban proposes incentives for broad-based employee equity (ESOP-like structures) to build wealth and close gaps. He cites research indicating ESOP participants earn more and experience lower turnover, framing equity as a practical path to shared upside.
- •People fall behind without assets; equity can change household trajectories
- •Incentivize companies to share equity broadly rather than only raising taxes
- •ESOPs correlate with higher earnings and lower turnover than peers/unions
- •Employee ownership as a scalable complement/alternative to direct transfers
Populism, trust, and results: what political strategy should optimize for
Cuban critiques progressive populism that mirrors Trump-style promises without clear implementation, while giving credit to long-horizon policy pushes that delivered outcomes (e.g., living wage advocacy, CFPB). His north star is pragmatic: choose candidates and strategies that can execute, communicate, and deliver measurable improvements.
- •Populist promises work because they exploit the trust divide
- •Credit for persistence and tangible wins (living wage, CFPB) vs. empty rhetoric
- •Candidate archetype should combine sales ability, tech fluency, and execution
- •Politics should be evaluated by results delivered, not slogans
Technology in government: efficiency, AI-enabled services, and new redistribution tools
Cuban argues government should use AI to reduce administrative cost and friction, then return savings to citizens (via checks, credits, or UBI-like mechanisms). He frames this as a message Democrats could sell effectively: more efficient government that directly increases take-home benefits.
- •Use AI to improve government efficiency and reduce costs
- •Share savings with citizens (checks/tax credits) to make benefits tangible
- •Robot taxes/UBI discussed as potential tools if automation displaces work
- •“Government as a service”: reduce paperwork and friction for citizens/businesses
AI in education: from cheating panic to redesigned learning
Cuban acknowledges students will use AI to cheat but argues schools can adapt by changing assignments to emphasize analysis, debate, and synthesis. He sees AI as a democratizing, always-on tutor—expanding access to customized learning for any student with a phone and internet.
- •Cheating is real; the response should be redesigning pedagogy
- •Use AI for research, then assess judgment and prioritization in class discussion
- •AI as a personalized tutor for languages, subjects, and curiosity-driven learning
- •Education democratization expands opportunity outside elite institutions
AI in healthcare: better care, but an IP and data “siloing” arms race
Cuban expects healthcare quality to improve as clinicians and patients use LLMs and evidence tools, even if total spending rises due to expensive cures. The central conflict becomes IP/data ownership: he predicts hospitals and researchers will silo proprietary findings and either build their own models or auction access to foundation-model providers.
- •Healthcare will improve qualitatively; spending may rise due to costly curative therapies
- •Clinicians already use AI tools; patients can pre-work issues with LLMs
- •Publishing/patenting becomes risky if models absorb IP without compensation
- •Future: many specialized models; hospitals/universities may run their own LLMs
Healthcare reform blueprint: eliminate PBMs/insurers, restore price transparency, restructure risk
Cuban gives a detailed critique of insurance/PBM incentives: opacity, risk-shifting to patients, pre-auth burdens, underpayment tactics, and vertical-integration accounting games (intercompany transfers). His alternative emphasizes transparent pricing, shopping tools, catastrophic reinsurance, and government-backed financing within benchmarked price parameters.
- •PBMs/insurers add opacity and avoid risk; Medicare Part D examples where cash price < copay
- •High deductibles + low ability to pay turns hospitals into ‘subprime lenders’
- •Insurers underpay contracted amounts; pre-auth delays monetize time value of money
- •Proposed system: no premiums/PBMs, transparent prices, shopping tools, reinsurance + government-guaranteed financing tied to benchmarks
AI-native careers and entrepreneurship: the new systems integrators for small business
Cuban advises young builders to be ‘AI every day’ and focus on applying AI to real business processes, especially in the millions of small companies that lack AI expertise. He frames the opportunity as customized, agent-driven ‘new SaaS’ where domain knowledge plus AI fluency creates immediate ROI.
- •Entry-level programming jobs compress; AI-native operators become valuable
- •Most opportunity is in small businesses (far more than large enterprises)
- •Agents can automate auditing/ops tasks (e.g., shipping invoice overcharges)
- •Future SaaS: customized agents, process redesign, scalable playbooks across industries
Building wealth and winning in business: patterns, ‘death wars,’ and remixing technologies
Cuban explains his career through competition, pattern recognition, and ‘remixing’ new tech into practical advantages—citing streaming, HD, the Mavs’ tech adoption, and Cost Plus Drugs. He looks for ‘death wars’ where competitors overspend to win, creating openings for differentiated sellers and suppliers.
- •Business as a sport: intellectual challenge and competitive drive
- •Innovation strategy: remix known tools into new markets (Jobs’ ‘everything’s a remix’)
- •Cost Plus Drugs as outsider strategy: avoid the entrenched PBM incentives
- •‘Death wars’ create big buyers and fast-moving markets—AI mirrors the TV wars
NBA economics and team-building under the new CBA: TV value, draft picks, and apron math
Cuban attributes NBA franchise valuation growth primarily to TV/streaming deal escalation, then pivots to roster construction constraints under the new CBA. He argues draft picks are more valuable now, the second apron is unusually punitive, and patience (OKC/San Antonio style) is increasingly advantaged—while also discussing star economics, defense/offense tradeoffs, and the Luka trade frustration.
- •Franchise value surge driven by linear/streaming TV bidding wars
- •Second apron restrictions change roster-building; fewer max players likely
- •Draft picks have become more valuable; patience beats ‘win now’ more often
- •Stars are undervalued relative to max; the hard part is paying non-generational max players
- •Fan allegiance shifting from teams to players affects attendance/viewership dynamics
Family-first priorities, unconventional investing, and making AI broadly popular
Cuban closes by explaining why he stepped back from the Mavs and Shark Tank: to spend time with his kids. He describes an email-driven, concentrated investing style focused on helping companies become profitable without constant fundraising, then argues Silicon Valley must communicate AI’s upside—especially in Middle America—through jobs, education, and accessible pathways like bootcamps.
- •Time allocation shift: prioritize family as kids near adulthood
- •Lean investing approach: direct deals, large ownership stakes, hands-on help to profitability
- •Enjoyment in building durable, cash-flowing brands (vs. fund/portfolio machinery)
- •AI adoption depends on communicating upside, not just disruption narratives
- •Engage beyond SF: bootcamps and outreach to widen participation and reduce backlash
