Glad to see coverage of our work! However, I wish the article (especially the title and the summary) would‘ve presented our approach more carefully / from a different angle - as to avoid the readers interpreting it as if we „confronted LLMs with pain“ or think we made them „experience“ anything.
Daria Zakharova
Scientific American reports on some of our attempts to make progress towards a reliable test for AI sentience. www.scientificamerican.com/article/coul...
A new study shows that large language models make trade-offs to avoid pain, with possible implications for future AI welfare
