The Bay Area’s animal welfare movement wants to recruit AI
Bay Area advocates are training LLMs with animal-centric data to shape future superintelligent systems.
In San Francisco, the organization Sentient Futures convened a unique summit where animal welfare advocates and AI researchers gathered to strategize on recruiting artificial intelligence for their cause. The event, held at the Mox coworking space, brought together individuals deeply influenced by effective altruism—a philosophy focused on maximizing global good—who believe that the impending arrival of Artificial General Intelligence (AGI) presents a pivotal opportunity. Their core thesis is that if superintelligent AI systems will eventually make most of society's critical decisions, it is imperative to shape those systems now to value animal lives and mitigate suffering on a massive scale.
A key presentation came from Jasmine Brazilek, co-founder of Compassion in Machine Learning, who demonstrated a benchmark she created to evaluate how Large Language Models (LLMs) reason about animal welfare. She urged researchers to train their models, such as GPT-4 or Claude, using synthetic documents that embed concern for non-human sentience. This technical approach aims to proactively influence the values of future AI. The movement prioritizes large-scale, utilitarian interventions—like promoting cultivated meat to reduce factory farming or considering the welfare of vast insect populations—over traditional local charity, betting that aligned AI could be the most powerful tool for flipping the game board on global animal suffering.
- Sentient Futures summit connects effective altruists and AI researchers to align future AGI with animal welfare values.
- Advocate Jasmine Brazilek created a benchmark to measure LLM reasoning on animal issues and pushes for training with synthetic data.
- Movement focuses on large-scale solutions like lab-grown meat, arguing future AI decisions must account for non-human suffering.
Why It Matters
This effort seeks to embed ethical values into foundational AI systems, potentially determining how future technology governs global welfare.