Anthropic, the AI safety company known for publicly worrying about existential risks from artificial intelligence, just hired an economist who argues that a 33 percent chance of human extinction is an acceptable trade-off for dramatic economic growth. The hire of Chad Jones, a longtime Stanford economics professor, highlights a tension at the heart of how the AI industry approaches catastrophic risk.

What Did the Economist Actually Calculate?

Jones published research that weighed the acceptable level of existential risk from AI against potential economic benefits. In his analysis, he modeled a scenario where humanity faced a one percent annual probability of extinction over 40 years. Using standard economic modeling, he calculated that the cumulative survival probability would be approximately 67 percent, meaning a 33 percent chance of human extinction.

His conclusion was striking: "In other words, with log utility it is optimal to take a 1 in 3 chance of ending human existence in exchange for a 2/3 chance of dramatically raising living standards by a factor of 55," Jones wrote in his paper. In plain terms, he argued that if AI could boost living standards by 55 times over, the mathematical case for accepting that extinction risk becomes defensible under standard economic frameworks.

Why Does This Hiring Decision Matter for AI Safety?

Anthropic has built its public reputation on being the AI company most concerned about safety and existential risk. The company regularly emphasizes its commitment to responsible AI development and has positioned itself as more cautious than competitors like OpenAI. Hiring an economist who has published work normalizing one-in-three extinction odds creates a credibility question.

The tension runs deeper than one hire. Anthropic's public messaging about AI risks serves a dual purpose: it genuinely reflects concerns about safety, but it also reinforces the idea that AI systems are powerful enough to pose existential threats. This framing, while perhaps accurate, also benefits Anthropic by elevating the perceived importance and capabilities of its technology and research.

How Do Experts and the Public View This Risk Calculation?

Jones's framework has drawn sharp criticism from those who argue that extinction risk cannot be treated like a standard financial decision. The logic of accepting a one-in-three extinction probability because of potential economic gains strikes many as fundamentally flawed.

• The Irreversibility Problem: Unlike financial losses that can be recovered over time, human extinction is permanent and cannot be offset by future economic gains, no matter how large.
• The Assumption of Survival: Jones's calculation assumes that if humanity survives the initial AI transition, it will enjoy the promised 55-fold increase in living standards, an assumption that may not hold if the survival margin is narrow.
• The Moral Hazard Issue: Accepting high extinction probabilities as mathematically justified could create incentives for AI developers to take greater risks, knowing that potential economic upside provides a rationalization.

One Reddit commenter captured the public skepticism bluntly: "This isn't poker where you let the math guide you because even if you lose now on a net positive decision it works out in the long run. Ending human existence requires a tad more discretion".

One Reddit

What Does This Reveal About AI Safety Culture?

The hiring raises questions about how seriously AI safety organizations weigh existential risk in practice versus in public messaging. Anthropic's commitment to safety is not necessarily undermined by hiring Jones, but the move suggests that the company may be more comfortable with risk frameworks that justify aggressive AI development than its public positioning implies.

This pattern extends beyond hiring decisions. Anthropic has faced criticism for its Claude AI model being used to select strike targets in Iran, despite the company's public disputes with the Pentagon over safe AI deployment. The gap between stated safety priorities and actual deployment decisions suggests that economic and strategic pressures may influence how safety concerns are weighed in practice.

Steps to Understanding AI Risk Frameworks

For those trying to make sense of how AI companies approach existential risk, several key concepts help clarify the debate:

• Expected Value Calculations: Economists like Jones use mathematical frameworks that multiply the probability of an outcome by its value. A small probability of a massive benefit can mathematically outweigh a large probability of a small loss, which is how extinction risk becomes "acceptable" in economic terms.
• Discount Rates and Time Horizons: Economic models often discount future benefits and risks based on how far away they are. This means that extinction risks decades away may be weighted less heavily than immediate economic gains, a practice that critics argue is inappropriate for existential threats.
• The Difference Between Risk and Uncertainty: Jones's framework assumes we can calculate probabilities precisely, but existential AI risk involves deep uncertainty where we may not even know what we don't know about how advanced AI systems will behave.

The hiring of Chad Jones at Anthropic illustrates a fundamental tension in AI safety: the same economic logic that justifies massive investment in AI development can also be used to justify accepting catastrophic risks. Whether that tension can be resolved depends on whether AI safety advocates can develop frameworks that treat extinction differently from ordinary financial decisions.