Draft:Optional Work Index

The Prediction

The index is anchored to a series of public statements made by Elon Musk, the chief executive of Tesla and xAI and founder of SpaceX,^[1] in which he has argued that the convergence of advanced artificial intelligence and humanoid robotics will render traditional employment economically optional within 10 to 20 years.^[2] In Musk's framing, work would transition into a "voluntary, hobby-like activity,"^[3] driven by humanoid robots and superintelligent AI, bringing what he has described as "universal high income"^[4] and sustainable abundance, rendering money "largely irrelevant."^[5] The baseline prediction horizon used by the index is 2034, with an outer horizon of 2044, anchored to January 2024 as the model's start date.

Musk has made several subsidiary predictions that are incorporated into the index as tracked milestones, including that Tesla's Optimus humanoid robot will be produced at a rate of one million units per year at Gigafactory Texas,^[6] that Optimus will be available for external commercial purchase at under $20,000,^[7] and that xAI's Grok language model will achieve frontier model leadership across reasoning benchmarks.^[8]

Motivation

The OWI was created in recognition that the question of whether AI and robotics will make work optional is one of the most consequential empirical questions of the 21st century. The index takes no position on whether the prediction will prove correct. Its function is to measure how fast the relevant conditions are developing, and to make that measurement transparent, reproducible, and publicly auditable.

The project has been described by its creator as sharing the curiosity and spirit of The Limits to Growth (Meadows et al., 1972)^[9] in its attempt to identify the behavioural tendencies and limits of the global system as it pertains to the future of human labour.

Sub-Indices

AI Capability (28% weight) measures the advancement of artificial intelligence systems across cognitive tasks previously requiring human workers. Data sources include performance on standardised benchmarks — SWE-bench Verified for real-world software engineering,^[10] GPQA Diamond for PhD-level scientific reasoning,^[11] ARC-AGI for general fluid reasoning,^[12] HumanEval for code generation,^[13] and Humanity's Last Exam for frontier scientific knowledge.^[14] The sub-index also incorporates data from Stanford HAI,^[15] the METR organisation,^[16] McKinsey Global Institute,^[17] and Goldman Sachs Global Investment Research.^[18]

Humanoid Robotics (22% weight) tracks the physical automation of labour through deployable robotic systems capable of performing human physical work. It draws on the International Federation of Robotics World Robotics Report,^[19] cost curve analysis from ARK Invest^[20] and Goldman Sachs,^[21] and FRED economic data series including the Industrial Production Manufacturing Index (IPMAN) and total manufacturing employment (MANEMP).^[22] The core signal is the divergence between industrial output and manufacturing employment: when output rises while jobs fall, the index treats the gap as evidence of robotic substitution of human physical labour.

Labor Market Shift (20% weight) monitors structural changes in employment patterns attributable to automation, distinguishing cyclical from structural displacement. Primary sources include the United States Bureau of Labor Statistics Job Openings and Labor Turnover Survey (JOLTS),^[23] the FRED series for job openings (JTSJOL),^[24] International Labour Organization employment data,^[25] OECD labour market statistics,^[26] and the World Economic Forum Future of Jobs Report.^[27]

Economic Abundance (18% weight) measures the degree to which AI-driven productivity gains are generating broadly accessible material wealth. It incorporates IMF World Economic Outlook GDP growth data,^[28] BLS productivity statistics,^[29] and universal basic income program data from GiveDirectly.^[30]

Wealth Distribution (12% weight) assesses whether the economic gains from automation are being distributed broadly or concentrated narrowly. Sources include WID.world global inequality data,^[31] World Bank extreme poverty metrics,^[32] and OECD income distribution statistics.^[33]

Scoring Formula

The composite OWI score is computed as:

OWI = Σ((SubIndex_i + LiveAdj_i) × W_i) × M + δ

Where:

• SubIndex_i is the current score for each of the five sub-indices
• LiveAdj_i is the real-time news signal adjustment for each category
• W_i is the category weight
• M is a milestone bonus multiplier, log-scaled from the count of confirmed historical events: min(8.0, log(n+1, 2) × 2.8)
• δ is a trajectory correction factor

News Signal

Each day, the OWI engine fetches current headlines via the NewsAPI.ai service^[34] and evaluates them using Claude, an AI language model developed by Anthropic,^[35] which scores each headline on a scale of −5 to +5 for its relevance and directional impact on the optional work prediction timeline. The model also identifies any headlines that may confirm a watched milestone target, flagging high-confidence detections for automatic promotion in the milestone database.

Milestone System

The OWI maintains a structured database of two classes of events: confirmed milestones and watched targets.

Confirmed milestones are real-world occurrences verified by a published source URL. Each confirmed event contributes to the milestone bonus component of the OWI score via a logarithmic scaling function. As of March 2026, the index has confirmed 83 historical milestones spanning the period from 2020 to 2026.

Watched targets are future events grounded in published research from McKinsey Global Institute,^[36] Goldman Sachs,^[37] the IMF,^[38] WEF,^[39] Oxford Martin School,^[40] IFR,^[41] ARK Invest,^[42] and other credible institutional sources. As of March 2026, the index tracks over 300 watched targets spanning all five categories and all years from 2026 to 2044.