Appendix III · Reading the Signals
Appendix III: How to Tell Which Stage We’re In
Every day the news shouts “disruption,” but between a dazzling demo and your household bills stand several doors. This page offers no forecast probabilities and no investment advice—just a measuring stick: any change must pass four gates in order—“it can be done,” “it can be done at scale,” “it can be done cheaply,” “everyone can get it”—before it truly changes ordinary lives.
To judge the stage, don’t look at the brightest light; look for the first door that’s still shut.
The Four-Gate Checker
Pick a change you care about—say, “driverless taxis” or “AI tutors”—and tick, in order, each gate for which you can find reliable evidence. The result shows where it first gets stuck.
Stuck before Gate 1: this is still an idea or a pitch. Worth watching, but not a basis for major life decisions.
Stuck at “deployment at scale”: it is technically feasible, but engineering, regulation and liability still stand between it and your life; the typical timescale is years.
Stuck at “falling cost”: it already runs in the real world, but only subsidized or premium users can afford it. This is the most common state of the AGI transition: replacement has begun, cheapness has not arrived.
Stuck at “universal access”: technology and cost are both ready, and only distribution and institutions decide whether lives change. This is exactly the fault line between the AGI transition and the ASI post-scarcity branch.
All four gates passed: basic needs in this domain are approaching “infrastructure.” Double-check that the evidence comes from independent sources, not a single vendor.
Signal Table for the Seven Domains
| Domain | Signals the AGI transition is accelerating | Signals approaching the ASI branch’s premises | Easily misread signals |
|---|---|---|---|
| 🌾 Food | AI adoption rises in farming and retail and waste falls, but corporate margins improve before retail prices do. | Nutritionally equivalent synthetic foods sell below conventional foods through mainstream channels, and energy costs keep falling. | A cultivated meat product “launches” or gets approved: that’s only Gate 1 or 2, not affordability or wide access. |
| 👕 Clothing | The number of styles and return rates rise together, and “AI-designed” becomes a marketing hook. | Automated sewing and closed-loop recycling reach mass production at everyday prices. | Smart-fabric fashion shows: mostly high-end showcases—check wash durability and price. |
| 🏠 Dwelling | Office vacancies rise, prime locations and ordinary homes diverge in price, and mortgage defaults overlap with regions of job replacement. | Autonomous construction is certified and scales, building costs keep falling sharply, and reforms to how housing rights are allocated begin. | “A house 3D-printed in 24 hours”: a fast structural shell does not mean total cost, permitting and utility connections fall at the same pace. |
| 🚗 Mobility | Cities and miles of driverless operation grow rapidly, and the number of professional drivers falls. | Accident and intervention rates stay significantly below human levels over time, all-in fares undercut private cars, and average commute times fall. | “Zero accidents” in a single-city pilot: small samples—look for long-term data across cities and weather. |
| 📚 Learning | AI tutoring spreads, entry-level hiring contracts, and in-person exams and portfolios carry more weight. | Employers widely drop degree requirements, and social protection is decoupled from employment. | A short-term test score gain: look for long-term transfer of learning and fairness. |
| 🧬 Longevity | More AI-discovered drug candidates enter clinical trials, care shortages widen, and debates over pension reform heat up. | Therapies for aging are approved on human clinical endpoints and priced within universal reach. | Lifespan extension in mice, or a billionaire’s personal anti-aging regimen: still far from population-level evidence. |
| ⚒️ Vocation | New-graduate hiring falls as a share, senior workers multiply their output, and labor’s share of income declines. | Social protection funded by taxes on capital or compute scales up, and the means of survival decouple from employment. | “How many new jobs AI has created”: look at job quality, income and how many people they cover. |
If You Are…: Watch Lists for Four Life Situations
Situation 01
A young person starting out
Watch the share of new-graduate hiring in your field, not the industry’s total output. Build work and relationships that real people can verify, and use roles that still offer an apprenticeship path, or require on-site judgment or accountability, as stepping stones.
Situation 02
A mid-career family with a mortgage
Put “how replaceable your income is” and “how long your debt runs” on the same sheet. The risk isn’t home prices themselves but the mismatch between future wages and monthly payments; watch what share of the tasks in your role is being automated, and how your company’s staffing is changing.
Situation 03
Parents
Degrees still matter, but they are shifting from “proof of ability” to “one proof among many.” Protect your children’s curiosity, focus and ability to collaborate, and keep an eye on changing admissions and hiring rules instead of betting on a single skill that “will never be replaced.”
Situation 04
Someone nearing retirement
Watch whether the funding sources for pensions and health insurance are being reformed, and whether care services are available. News of medical breakthroughs is exciting, but plan your life around therapies that are approved and affordable.
The above is a framework for observation, not investment, career, medical or legal advice.
Research Entry Points
- Eloundou et al., GPTs are GPTs (2023): how exposed different occupations’ tasks are to large language models.
- Brynjolfsson, Li and Raymond, Generative AI at Work (NBER, 2023): productivity changes from generative AI in a real workplace, with the largest gains for less experienced workers.
- Acemoglu and Restrepo, Automation and New Tasks (NBER, 2019): the twin effects of automation replacing tasks and creating new ones.
- FAO et al., The State of Food Security and Nutrition in the World (SOFI): annual data on hunger and the affordability of food.
- World Health Organization, Global Health Estimates: life expectancy and healthy life expectancy.
- Waymo, Safety: driverless miles and crash comparison data.
- Nick Bostrom, What is a Singleton?: the source of the concept of a unified governance structure.
These links are starting points for further reading; citing them does not mean this appendix’s scenario conclusions have been proven.
“The future does not arrive all at once. It appears first in demos, then in the lives of the rich, and finally—if we get distribution right—in everyone’s life.”