Interview
a16z Podcast | Companies, Networks, Crowds
- Authors and Context: Eric Brynjolfsson and Andrew "Andy" McAfee (MIT) discuss their third collaborative book, Machine, Platform, Crowd, following Race Against the Machine and The Second Machine Age.
- Income Stagnation vs. Tech Progress: While median income stagnated over the last decade despite technological advancement, the authors argue technology increases the total "pie" but does not guarantee equitable distribution; the top 1% captured disproportionate gains while others were left behind.
- Fundamental Economic Primitives: The conversation establishes three core building blocks:
- Network Effects (Demand-side economies of scale): Value increases as more users join (e.g., WhatsApp, Facebook).
- Economies of Scale (Supply-side): Costs decrease as production volume increases.
- Complements: Products that increase value when used together (e.g., razors/blades, iPhone/apps), capable of shifting demand curves even without price changes for the primary good.
- Apple App Store Case Study: Steve Jobs initially resisted opening the App Store but caved due to pressure from board members and executives who recognized that external developers created a massive ecosystem of complements that drove iPhone demand.
- Result: Opening the store unleashed a "tidal wave" of complimentary goods (including free apps like Angry Birds) that nudged demand upward for the hardware.
- Future of the Firm: The authors argue firms will not be replaced by decentralized networks (e.g., blockchain, DAOs) despite advances in technology.
- Incomplete Contracts Theory: It is impossible to write a contract specifying every contingency; ownership matters because it grants "residual rights of control" over unspecified future events.
- DAO Failure: The 2016 "The DAO" experiment failed when a hack required a centralized, autocratic decision to reset the system, proving the necessity of ownership structures for conflict resolution.
- Bounded Rationality: Even with advanced AI, human complexity and the "Red Queen" phenomenon (where competitors adapt equally fast) prevent perfect central planning or prediction.
- Crowdsourcing and Joy's Law:
- Definition: Joy's Law states that most smart people work for competitors; digital connectivity allows firms to tap into external talent pools.
- NIH Genome Sequencing Case: A crowdsourced algorithmic challenge reduced sequencing time from 4 hours to 10 seconds and accuracy from 75% to 80%.
- Surprise: The top solvers had no background in life sciences, highlighting the value of diversity and cross-domain thinking.
- Crowd vs. Core: Successful firms must rebalance resources, moving from a closed "core" to interfaces that harness the "crowd" for innovation.
- Human-AI Symbiosis:
- Problem Definition: Machines excel at providing answers, but humans remain superior at defining problems and framing context (a "native speaker" advantage in a human-created world).
- Augmentation Models: The most effective systems combine human and machine capabilities:
- Sales: Udacity used AI to prompt human sales reps with successful reply scripts during common queries, leaving complex issues to humans.
- Healthcare: AI serves as the diagnostic expert (analyzing lab results and images), while humans provide empathy, context, and treatment adherence strategies.
- Organizational Strategy:
- Rejection of "One-Size-Fits-All": Success depends on the specific combination of machines, platforms, and crowds; there is no single recipe.
- Overweighting the Core: Many firms fail by spending too much managerial bandwidth on existing core capabilities rather than integrating external crowds.
- Open Source Strategy: Modern enterprise software leverages open-source communities for legitimacy and development while the core firm defines the problem and captures value.
- Innovation Antibodies: Large organizations naturally resist external innovation ("not invented here" syndrome).
- Mitigation: To absorb external innovation, firms must isolate new ventures or deploy them in forward-thinking departments where the problem is clearly defined and leadership is open to change.
- Economic Trends: The authors express concern over a decline in startup formation and young firms in the US, noting that dynamic economies require a constant mix of established firms and new entrants.