Eliezer Yudkowsky

Overview

Eliezer Yudkowsky is presented as an AI-safety writer and foundational internet-rationalist figure.

Current Profile

The episode connects rationalist belief revision with AI alignment through Yudkowsky. It says he moved from favoring faster development of powerful AI toward warning about superintelligence after judging alignment much harder than expected. Value fragility and the orthogonality of intelligence and goals summarize that risk frame.

Key Characteristics

• Foundational writer in online rationalist culture.
• Early contributor to AI alignment and catastrophic-risk discourse.
• Presented as changing position after revising his view of alignment difficulty.
• Emphasizes that intelligence does not guarantee benevolent goals.

Evidence

• Role: E257｜对话Vitalik：AI安全背后的哲学思想，有效利他主义如何影响硅谷？ identifies Yudkowsky as a rationalist and AI-safety founder figure.
• Revision: E257｜对话Vitalik：AI安全背后的哲学思想，有效利他主义如何影响硅谷？ attributes his move from acceleration toward warning to alignment difficulty.
• Concepts: E257｜对话Vitalik：AI安全背后的哲学思想，有效利他主义如何影响硅谷？ explains value fragility and intelligence-goal orthogonality through simplified examples.

Qualifications

The summary compresses a long intellectual history and does not establish the technical validity of his forecasts. Its examples are explanatory, not a complete alignment theory.

What Changed

• Created the profile around rationalist culture, belief revision, and early alignment risk.

Relationships

• Rationalist Community - culture in which he is presented as foundational.
• AI Alignment Governance - broader institutional problem beyond technical risk.
• Existential Risk Probability Estimation - adjacent decision practice.
• Effective Altruism - overlapping but independently rooted movement.