Home

Check out my YouTube channel – please like, subscribe, comment and share!

This image has an empty alt attribute; its file name is ab6765630000ba8a60852c713be5bfdcb3c67f08.jpg

Safe As on YouTube

If you’d like to support my quest to share research locked behind paywalls, then please feel free to buy me a coffee

Latest from the Blog

High Reliability Organizations and Healthcare Outcomes on Patients and Staff: Scoping Review

Do HRO (High Reliability Organisation) principles improve healthcare patient & staff outcomes? This scoping study evaluated 9 studies. (Note – still under peer review, so use with caution). Extracts: ·        “HRO implementation was associated with fewer adverse events, positive staff perceptions on patient safety, and improved psychological safety” ·        Studies used a mixture of quantitative and qualitative…

Did we ‘train’ AI LLMs to hallucinate?

Did we ‘train’ AI LLMs to hallucinate? This study from OpenAI argues that hallucinations “need not be mysterious—they originate simply as errors in binary classification”. And, interestingly, hallucinations, persist, in part because the models “are optimized to be good test-takers, and guessing when uncertain improves test performance”. That is, they’re rewarded in training for guessing…

Part 1 interview with Sidney Dekker now live – check it out

Part 1 interview with Sidney Dekker is now live. We talk zero harm, behavioural safety, restorative culture and more. PS. You don’t see it on the screen – but there’s a timer running. Part 2 released sometime soon. If you find it interesting, then please hit like and leave a comment directly on YouTube, and…

Get new content delivered directly to your inbox.