#33 - Dr Anders Sandberg on what if we ended ageing, solar flares & the annual risk of nuclear war

Life Extension Prioritization

• Anders Sandberg identifies aging as a neglected area of research with "tremendous scope," responsible for nearly half of lost life years.
• He argues that while individual cryonics may be a valid personal priority (valuing >$7M life expectancy), it is likely less cost-effective per life-year saved than funding biomedical research into slowing aging.
• Sandberg suggests redirecting a small percentage (potentially 10%) of health budgets toward root-cause aging research, anticipating a "longevity dividend" where reduced chronic disease lowers societal costs and increases productivity.
• He notes that current health systems over-invest in treating symptoms of age-related diseases (cancer, diabetes) rather than the underlying aging process, which could yield higher long-term payoffs.

Societal Risks and Solutions of Long Lifespans

• Political Leadership: Statistical modeling suggests non-aging dictators would stay in power approximately 33% longer (increasing from an average of 12 years to 16), though most lose power due to internal coups rather than biological aging.
• Institutional Solutions: To prevent "grey blocks" or institutional sclerosis, Sandberg proposes term limits for high-level jobs (e.g., capping tenure at a century) to ensure generational turnover.
• Cognitive Flexibility: Contrary to the fear that long-lived populations become sclerotic, he cites UK data showing political views can become more flexible after retirement, suggesting culture and institutional design matter more than biological age.
• Intergenerational Equity: A major concern is the "grey block" scenario where older generations hold wealth and influence, potentially stifling youth innovation; he advocates for new systems to facilitate wealth and idea transfer.

Feasibility of Life Extension

• Scientific Progress: Since the 1990s, the scientific consensus shifted from "aging is immutable" to "aging is modifiable," driven by the discovery of genetic interventions in worms and mice (e.g., Cynthia Kenyon's work).
• Current Interventions: The anti-diabetic drug metformin shows promise in extending life in animal models and reducing age-related diseases in humans; Sandberg plans to begin taking it in his 40s.
• Risk Tolerance: Sandberg argues that as individuals approach higher mortality rates, they should consider experimental interventions (like rapamycin or senolytics) where the risk of the treatment is outweighed by the certainty of death from aging.
• Data Collection: He advocates for "observational trials" where individuals self-experiment with enhancements but share data openly to accelerate knowledge, criticizing current risk-minimization regimes that stifle learning.

Anthropic Shadow and Nuclear Risk

• Anthropic Bias: Sandberg explains that we cannot observe civilizations that have gone extinct, creating an "anthropic shadow" that biases our view of risk; a history of near-misses might not indicate safety, but rather that dangerous outcomes preclude observers.
• Risk Estimates: Adjusting for this bias, the annual probability of nuclear war is estimated between 0.1% and 1% per year, which he characterizes as "disturbingly large."
• Near-Miss Analysis: Analysis of near-miss events (e.g., Stanislav Petrov, 1983) suggests the "anthropic selection effect" is mild; the frequency of near-misses does not strongly correlate with a low-risk world, implying the nuclear system remains fragile.

The Unilateralist Curse

• Definition: The "unilateralist curse" describes a scenario where multiple uncoordinated agents independently decide to take high-risk actions (e.g., releasing a GMO, geoengineering) because they believe their specific action is justified.
• Probability Logic: As the number of potential agents increases, the probability that at least one person will act on a flawed premise rises, even if all agents attempt to be rational.
• Prescribed Behavior: Individuals should adopt a principle of conformity, acting more conservatively than they would in isolation, discounting their own judgment by the likelihood that others in similar situations would make the same mistake.

Natural and Systemic Catastrophic Risks

• Solar Flares: A Carrington-class solar event could induce currents that destroy global power grid transformers; the average return time is estimated at 150 years, and supply chain delays (months to replace transformers) pose a systemic risk.
• Supply Chain Vulnerability: Society relies on "linchpin" products with low global redundancy (e.g., a single Japanese factory producing a specialized resin for memory chips); the 1990s fire at this facility caused a global chip shortage.
• Other Natural Risks: Gamma-ray bursts and nearby supernovas are statistically unlikely but theoretically capable of sterilizing biospheres; ice ages and climate shifts are slow-moving risks with potential for major geographic disruption.

Ethics of Sex Robots and Automation

• Counter-Argument: Sandberg challenges the campaign against sex robots, arguing that banning them is a weak strategy compared to addressing the root cause (objectification).
• Comparison to Killer Robots: Unlike killer robots, which directly cause harm (death), sex robots are a technology of pleasure; banning them removes the activity without necessarily solving the societal issue.
• Supernormal Stimuli: A more significant risk is the development of "persuasion robots" or superstimuli (exaggerated features in pornography/robots) that could lead to addiction and distorted human expectations, potentially harming reproduction and social cohesion.
• Technological Adaptation: He posits that human culture adapts to new technologies (printing, internet) and that banning specific tools may not prevent the underlying behavioral shifts.

Sleep Minimization and Enhancement

• Value of Sleep: Sleep consumes one-third of human life; improving sleep efficiency could add years to a person's waking life.
• Biological Uncertainty: The exact function of sleep is unknown; theories regarding memory consolidation do not explain why total sleep deprivation causes death in lab animals.
• Enhancement Risks: While sleep reduction seems beneficial for productivity, evolution likely preserves sleep due to critical trade-offs; attempting to eliminate sleep could be dangerous, though improving sleep quality (treating insomnia) is highly prioritized.
• Polyphasic Sleep: He notes that polyphasic sleep attempts often fail due to incompatibility with social structures and the biological need for consolidated rest.

Forecasting and Methodology

• Calibration: Sandberg engages in annual forecasting to calibrate his probability estimates, finding he tends to be overconfident about highly probable events.
• Problem Definition: He highlights the difficulty in formulating unambiguous predictions, noting that specific dates for complex events (e.g., "human-level AI by 202X") are often misleadingly vague.
• Advice for Researchers: He recommends interdisciplinary breadth, urging future researchers to read introductory textbooks on diverse fields (e.g., solid-state mechanics, economics, philosophy) to identify connections and solve complex problems.