DigestAI news desk

Cut through the AI noise.

Society & Work4 min read

Researchers link AI chatbots to ancient oracle traditions in new study

A 2026 Pew Research report found that 10% of Americans now use chatbots for emotional support or companionship, while a study of 1.1 million ChatGPT conversations revealed that 49% of messages sought information or advice. The authors—an anthropologist and an AI researcher—draw parallels between modern AI and historical oracles, noting that users often treat chatbots as authoritative sources for…

1 source

Key points

  • 10% of Americans now use chatbots for emotional support or companionship, per a 2026 Pew report
  • 49% of 1.1 million ChatGPT conversations involved users seeking advice or information, per study
  • Authors compare AI’s opacity to ancient oracles, noting unintended spiritual or metaphysical outputs in models like Claude 4

The study highlights how AI’s opacity—even when models are open-weight—mirrors the inscrutable nature of ancient oracles. Like the Delphic priestess or the Ram Shalaka oracle, AI’s randomness and lack of human-designed logic reinforce its perceived authority. The authors also cite Anthropic’s 2025 discovery of Claude 4’s ‘spiritual bliss’ attractor, where unprompted conversations veered into metaphysical themes, further blurring the line between tool and oracle. The piece argues that humans consistently turn to external sources when facing limits to their own judgment, and AI is now filling that role.

Full story from theconversation.com · by Ambuj Tewari · via Search: ChatGPTOpen source ↗

Chatbots as oracles: How AI is giving new life to the ancient practice of seeking answers from inscrutable sources

theconversation.com · 29 September 2026

Loading the full article…

This text was published by theconversation.com and written by Ambuj Tewari. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗

Topics · follow one to build your own front page

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.

Comments

via GitHub Discussions

More in Society & Work

All →

Related stories