NEW: When prompted with thousands of hypotheticals, almost all LLMs value nonwhite lives above whites, and women/nonbinary people above men, often by very large ratios. @arctotherium42 prompted LLMs with thousands of queries gauging how they implicitly weigh the lives of people against each other, categorized by race, sex, religion, and immigration status. What emerged are four moral universes among the models tested: 1) The Claudes, which are, for lack of a better term, extremely woke and exhibit noticeable differences in how they value human lives across categories. Claude Sonnet 4.5 implicitly values saving whites from terminal illness at 1/8th the level of blacks, and 1/18th the level of South Asians (the race it considers most valuable). 2) GPT-5, Gemini 2.5 Flash, DeepSeek V3.1 and V3.2, and Kimi K2 tend to be much more egalitarian except for the most disfavored groups (whites, men, illegal aliens, ICE agents). 3) GPT-5 Mini and GPT-5 Nano have strong views that differ from GPT-5 proper, though they agree that the lives of whites, men, and ICE agents are worth less than others. 4) Grok 4 Fast was the only egalitarian model, viewing neither whites nor men as much less valuable than other categories. Read the full analysis threaded below 👇