Tech companies seek religious guidance to make AI morally good

Summarised from 2 outlets · Updated 25 Oct, 11:31 · Archive

Tech companies are consulting religious and faith leaders to help develop morally sound artificial intelligence systems.

Tech companies developing artificial intelligence are increasingly turning to religious leaders and philosophers for guidance on embedding moral values into their systems. Anthropic co-founder Chris Olah told the Vatican earlier this year that questions of teaching chatbots right from wrong cannot be answered by laboratories alone, but are matters that religious traditions have grappled with for centuries.

Companies like OpenAI and Anthropic are seeking to make their AI models reflect broad moral principles drawn from multiple religious and philosophical traditions. OpenAI CEO Sam Altman stated last year that ChatGPT should embody 'a weighted average' of humanity's moral views. Because tech companies are not tied to any single religion, their approach aims to aggregate moral ideas across different traditions.

However, experts and adherents of various faiths warn that this strategy fails to account for fundamental contradictions between different worldviews and religious traditions. The companies have received little public scrutiny regarding what they actually understand religion to be or what they hope to gain from consulting faith leaders, even as executives publicly highlight concerns about AI's potential for harm.

How it is being reported

• To make AI morally good, tech companies are turning to religion. But some say it isn't that simpleTech companies are increasingly turning to religion for guidance on developing ethical AI models.Clickorlando · 25 Oct, 10:57
• To make AI morally good, tech companies are turning to religion. But some say it isn't that simpleTech companies are increasingly turning to religion for guidance on developing ethical AI models.News4Jax · 25 Oct, 10:57

In this story: Anthropic · Chris Olah · OpenAI · Sam Altman · Vatican

This summary was written by AI from the headlines and standfirsts above, and states as fact only what at least two outlets report. How we use AI · Report a problem