The author proposes a "fun formula" to describe preferences for a future with aligned superintelligence, focusing on maximizing a "fun field" across sentient minds. Three potential aggregation methods are discussed: summing total fun across all beings and time, measuring the variety of distinct fun experiences realized, and maximizing the rate of fun's growth. The author expresses a personal inclination towards maximizing variety, likening it to filling a "glass" of possible good experiences. AI
IMPACT Explores philosophical frameworks for evaluating AI futures, influencing long-term alignment goals.
RANK_REASON The cluster contains an opinion piece discussing philosophical preferences for a future with superintelligence, rather than a concrete release or event.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →