Too Much Trust in AI Poses Unexpected Threats to the Scientific Process

It’s vital to “keep humans in the loop” to avoid humanizing machine-learning models in research

https://www.scientificamerican.com/article/trust-ai-science-risks/