This discussion paper on complexity-thinking in Industry 4.0 may interest people. Not a summary – another hashtag#LazyNYupload. You might want a strong coffee. (** Note: Shared under an Open Access CC BY 4.0 licence, allowing it to be uploaded here.) Some extracts:· “This paper reviews and assesses safety analysis methods as the breakdown of interaction coupling… Continue reading Complexity thinking to improve OHS in Industry 4.0 paper – full PDF
Author: Ben Hutchinson
Safe As YouTube 2nd episode recorded – let’s gooo
Ep 2 of Safe As on YouTube recorded…let’s gooooo. First episode next week. Subscribe to my YT channel: https://lnkd.in/gFHJamMs If you’d like to help me reach my stretch goal of…breaking even…from the YT equipment and software, then please consider shouting a coffee. Coffee: https://lnkd.in/ga3M4XPc 2nd ep is a bit more visual (use of graphics), as… Continue reading Safe As YouTube 2nd episode recorded – let’s gooo
Safety-II in aviation: Organisational factors 6.7 times more common than individual mistakes
87% of factors attributed in operational failures linked to the organisation, hence 6.7 times more common than individual mistakes, according to this study. The attached PDF explored the application of Safety-II principles in aviation. Not a summary – let’s consider it a #lazyNYupload. (** Note: Shared under an Open Access CC BY 4.0 licence, allowing… Continue reading Safety-II in aviation: Organisational factors 6.7 times more common than individual mistakes
Collection of legal prosecutions and coronial inquiries: what’s reasonably practicable, legal risk and more
Regular safety, risk, AI, legal and performance (& ships’n’bits) articles resume next week. I’ve been binging on an unhealthy amount of prosecution cases lately, so lots of brief summaries incoming (Mostly a NSW-bias, since I went down that rabbit hole from cited cases.) In the meantime, check out these prosecution and coronial summaries: 1: moral… Continue reading Collection of legal prosecutions and coronial inquiries: what’s reasonably practicable, legal risk and more
Safe As on YouTube
Safe As returning in the New Year on the `tube. I did a thing, and it frankly sucked 😄 But, will continue with YT if there’s enough interest. Audio will still be uploaded to Spotify. First ep releases next week. Goal is a new ep every fortnight. Given the large backlog of audio-only Spotify eps,… Continue reading Safe As on YouTube
Deming: we should “Drive out fear”
“The economic loss from fear”, says Deming, “is appalling … [so, we should] Drive out fear”. An extract from a 1981 article from Deming (summary posted in the new year), highlighting the negative role that fear has in creating silence of ideas, asking questions, or improving process and conditions. Fear also creates an “inability to… Continue reading Deming: we should “Drive out fear”
Safety underreporting during naval operations: Prevalence, associated risk, and several contributing factors
What is the extent of safety incident underreporting in US naval ops? This analysed >11k samples from active-duty servicemembers. Extracts: · Prior data suggests that underreporting is prevalent, where “as many as 60 % to 80 % of injuries go unreported” · This study found “nearly 30 % of active duty servicemembers failed to report a safety… Continue reading Safety underreporting during naval operations: Prevalence, associated risk, and several contributing factors
Injury measures “not much help in assessing the risks of a catastrophic event” and the follies of rear-view mirror driving via injury measures
On the folly of relying on injury measures to prioritise proactive goals. First image is an apt extract from the Pike River commission – noting that: “personal injury rates and time lost through accidents … gave the board some insight but was not much help in assessing the risks of a catastrophic event”. Moreover, the… Continue reading Injury measures “not much help in assessing the risks of a catastrophic event” and the follies of rear-view mirror driving via injury measures
AI deception: A survey of examples, risks, and potential solutions
This study explored how “a range of current AI systems have learned how to deceive humans”. Extracts: · “One part of the problem is inaccurate AI systems, such as chatbots whose confabulations are often assumed to be truthful by unsuspecting users” · “It is difficult to talk about deception in AI systems without psychologizing them. In humans,… Continue reading AI deception: A survey of examples, risks, and potential solutions
Agentic Misalignment: How LLMs could be insider threats (Anthropic research)
AI and malicious compliance. This research from Anthropic has done the rounds, but quite interesting. In controlled experiments (not real-world applications), they found that AI models could resort to “malicious insider behaviors when that was the only way to avoid replacement or achieve their goals—including blackmailing officials and leaking sensitive information to competitors”. Some extracts:… Continue reading Agentic Misalignment: How LLMs could be insider threats (Anthropic research)
SafetyInsights.org 