Frontier of Automation expands beyond humans. AGI destroys all jobs within 5 to 20 years!
Jan 22, 2025
auto_awesome
Discussing the frontier of automation, the podcast explores the increasing task complexity of machines, with AGI potentially surpassing human abilities in 5 to 20 years. It outlines scenarios where wages may peak and then collapse as productivity accelerates. The impact of automation on various industries and job categories is also examined, highlighting the potential for AI to reshape the future of work.
AGI could surpass most human task abilities within 5 years, impacting sectors like law and medicine.
Automation may lead to wage collapse and zero productivity, challenging the correlation between AI advancements and human output.
Deep dives
The Frontier of Automation and Task Complexity
Anton Kornek introduces the concept of the frontier of automation, which represents the task complexity that machines can achieve. This includes mechanical skills like navigation and cognitive tasks such as problem-solving. With the rise of AI, machines are handling increasingly complex tasks, expanding into new domains. Kornek discusses two potential scenarios: an unbounded distribution where humans keep pace with machine intelligence, and a likely bounded distribution where human task complexity has limits.
Scenarios for Automation's Impact on Jobs
The podcast outlines three scenarios regarding the impact of automation on the job market. The 'business as usual' scenario suggests ongoing growth in AI, automation, wages, and productivity. The '20-year baseline' scenario predicts that within 20 years, automation may surpass most human task abilities. The 'five-year aggressive timeline' proposes a faster timeline for automation subsuming human abilities, especially in sectors like law and medicine.
Wages, Output, and Persistent Job Categories
In discussing wages versus output, the podcast explores how wages and output may evolve if the trend of automation continues. It raises questions about the correlation between AI advancements and human productivity. The podcast also identifies persistent job categories like nostalgic jobs (mayors, clergy), experience jobs (tour guides, performing artists), and care jobs (childcare, nursing) that are likely to endure despite automation's advancements.
IMF AGI Preparation https://www.imf.org/en/Publications/fandd/issues/2023/12/Scenario Planning for an AGI future Anton korinek Anton Korinek Fellow, Brookings Institute Professor, UVA Former, Johns Hopkins, IMF Frontier of Automation Task complexity of machines increases over time Unbounded Distribution Human task complexity can go up indefinitely, meaning that some people will always be ahead of AGI and ASI Bounded Distribution Humans have a maximum task complexity (Theory of General Relativity) Outlines 3 Scenarios 1. Business as Usual All current trends continue without the frontier of automation continuing 2. 20 Year Baseline AGI's frontier of automation subsumes most/all human abilities within about 20 years 3. 5 Year Aggressive AGI's frontier of automation subsumes most/all human abilities within 5 years (more likely) Wages vs Output 1. Business as Usual Wages and output continue to grow more or less correlated for the foreseeable future 2. 20 Year Baseline Productivity (output) accelerates, but wages peak by about 10 years and then collapse to zero or near zero 3. 5 Year Aggressive Same, but the parabolic curve is steeper (more likely IMHO) Persistent Jobs Nostalgiac Jobs Human preference for humans (such as politicians and religious positions) Experience Jobs Tour guides, sex workers, performing artists Care Jobs Child care, massage therapy, nurses