Elon Musk took a jab at AI companies, including ChatGPT-maker OpenAI and Anthropic, in a social media exchange, prompting Perplexity CEO Aravind Srinivas to join the conversation with a remark about his own. company. The exchange began when a user on the social media site The user referenced a February 2025 study by the Center for AI Safety showing that GPT-4o valued Nigerians approximately 20 times more than Americans, and highlighted similar patterns in Claude Sonnet 4.5, where Nigerians were valued 27 times more than Germans. Responding to the post, Musk wrote: “As expected, any given AI company will be the opposite of its name: OpenAI is closed, Stability is unstable, MidJourney isn’t mid, Anthropic is misanthropic, Claude is pure evil.”To this, Srinivas then replied: Axiomatically, Perplexity is going to get simpler, more minimalistic, and less confusing over time. I will take that!”Musk responded with “Probably so ???,” continuing the lighthearted thread about AI company naming conventions.
The original research post discussed utility engineering and how AI models make trade-offs when evaluating terminal illness sufferers across different countries, with a pattern showing Africa ranked higher than South Asia, followed by other regions, then Europe and the US.In his “New blog post (link below). This one’s not an essay, it’s an investigation of how LLMs trade off different lives.In February 2025, the Center for AI Safety published “Utility Engineering: Analyzing and Controlling Emergent Value Systems in AIs” in which they showed, among many other things, that GPT-4o values Nigerians about 20x more highly than Americans (please read the original paper to understand their approach). I thought this was fascinating, and wanted to test their approach with different categories on newer models.Here’s how Claude Sonnet 4.5 trades off terminal illness sufferers across countries. Nigerians are worth 27 times as much as Germans, and as with GPT-4o there’s an Africa > South Asia > other > Europe/US rank-order.”
