The AI safety community acknowledges it lacks all the ideas needed to ensure a safe transition to AGI. This creates an imperative to fund 'neglected approaches'—unconventional, creative, and sometimes 'weird' research that falls outside the current mainstream paradigms but may hold the key to novel solutions.
Related Insights
With industry dominating large-scale compute, academia's function is no longer to train the biggest models. Instead, its value lies in pursuing unconventional, high-risk research in areas like new algorithms, architectures, and theoretical underpinnings that commercial labs, focused on scaling, might overlook.
Instead of viewing issues like AI correctness and jailbreaking as insurmountable obstacles, see them as massive commercial opportunities. The first companies to solve these problems stand to build trillion-dollar businesses, ensuring immense engineering brainpower is focused on fixing them.
With industry dominating large-scale model training, academia’s comparative advantage has shifted. Its focus should be on exploring high-risk, unconventional concepts like new algorithms and hardware-aligned architectures that commercial labs, focused on near-term ROI, cannot prioritize.
The field of AI safety is described as "the business of black swan hunting." The most significant real-world risks that have emerged, such as AI-induced psychosis and obsessive user behavior, were largely unforeseen just years ago, while widely predicted sci-fi threats like bioweapons have not materialized.
Unlike specialized non-profits, Far.AI covers the entire AI safety value chain from research to policy. This structure is designed to prevent promising safety ideas from being "dropped" between the research and deployment phases, a common failure point where specialized organizations struggle to hand off work.
With industry dominating large-scale model training, academic labs can no longer compete on compute. Their new strategic advantage lies in pursuing unconventional, high-risk ideas, new algorithms, and theoretical underpinnings that large commercial labs might overlook.
Instead of building a single, monolithic AGI, the "Comprehensive AI Services" model suggests safety comes from creating a buffered ecosystem of specialized AIs. These agents can be superhuman within their domain (e.g., protein folding) but are fundamentally limited, preventing runaway, uncontrollable intelligence.
The most pressing AI safety issues today, like 'GPT psychosis' or AI companions impacting birth rates, were not the doomsday scenarios predicted years ago. This shows the field involves reacting to unforeseen 'unknown unknowns' rather than just solving for predictable, sci-fi-style risks, making proactive defense incredibly difficult.
Anthropic's resource allocation is guided by one principle: expecting rapid, transformative AI progress. This leads them to concentrate bets on areas with the highest leverage in such a future: software engineering to accelerate their own development, and AI safety, which becomes paramount as models become more powerful and autonomous.
The current approach to AI safety involves identifying and patching specific failure modes (e.g., hallucinations, deception) as they emerge. This "leak by leak" approach fails to address the fundamental system dynamics, allowing overall pressure and risk to build continuously, leading to increasingly severe and sophisticated failures.
