newsfilter.io
Interview

a16z Podcast | Automation, Jobs, & the Future of Work (and Income)

  • Tom Davenport (Babson University, MIT fellow) and Julia Kirby (Harvard University Press editor, HBR contributing editor) are co-authors of the upcoming book Only Humans Need Apply: Winners and Losers in the Age of Smart Machines.
  • The current AI landscape is described as the most "spring-like" period in history, characterized by high interest and widespread firm adoption compared to previous "AI winters."
  • Automation has silently permeated daily life for years, including invisible processes in banking (ATMs), airports, and early automated decision-making in insurance and credit underwriting a decade ago.
  • A three-wave model of automation is identified:
    • Wave 1: Machines automated dangerous physical work.
    • Wave 2: Computers automated dull, repetitive tasks (e.g., transcription).
    • Wave 3: Cognitive technologies now automate decision-making, eroding the perceived safety of knowledge work and advanced degrees.
  • The authors reject the notion of a "higher ground" immune to automation; instead, they propose augmentation (humans and machines as colleagues) as the primary future model over total replacement.
  • Automation targets specific tasks within jobs rather than entire roles, specifically removing codifiable, rules-based, and low-ambiguity work.
  • The Oxford University study cited estimates 47% of U.S. jobs are automatable, though the authors clarify this refers to tasks, not whole jobs, and lacks a timeline.
  • Five specific strategies for human adaptation ("stepping") in an augmented workplace are defined:
    • Stepping In: Monitoring and improving machine performance, such as underwriters correcting AI underwriting rules.
    • Stepping Up: Managerial roles that oversee automated systems, deciding when to adjust or deactivate algorithms (e.g., hedge fund managers).
    • Stepping Forward: Developing, marketing, and supporting intelligent technologies (e.g., IBM hiring for cognitive capabilities).
    • Stepping Aside: Focusing on uniquely human traits like creativity, complex communication, ambiguity handling, humor, and "taste" where human empathy is required (e.g., financial advisors acting as psychiatrists for clients).
    • Stepping Narrowly: Operating in niche areas (e.g., early-stage scientific discovery) where no economic case for automation exists yet.
  • Machine creativity (e.g., art, humor) currently lags behind human capability and often produces clichés; the authors suggest a future preference for human-crafted "artisan" work may emerge.
  • The authors explicitly reject the necessity of Universal Basic Income (UBI) as a solution to automation, arguing it relies on the false premise that humans will have "nothing to do."
  • They argue work is essential for human identity, structure, and satisfaction, citing that UBI experiments often lead to increased leisure (e.g., watching TV) rather than meaningful activity.
  • The proposed alternative to UBI is guaranteed work (compensated employment) rather than guaranteed income, to maintain incentive alignment.
  • A Swiss vote on a guaranteed basic income of roughly $2,500 is expected within a few weeks to serve as a large-scale experiment.
  • Organizations may reverse outsourcing trends, bringing work in-house because augmentation allows for greater innovation and margin retention compared to the "race to the bottom" cost-cutting of pure automation.
  • Management science is shifting from engineering-focused efficiency models to psychology-driven models that leverage human emotional intelligence and "human" competitive advantages.
  • The authors predict that organizational culture must explicitly state "augmentation is our objective" to prevent layoffs and liberate employees from tedious tasks.