The illusion of safety
Thread Starter
Joined: Jul 2003
Posts: 1,323
Likes: 54
From: An Island Province
The illusion of safety
The book "The illusion of safety" has been widely referenced and a ChatGPT review published:
https://chatgpt.com/share/687e3d50-9...8-b51a51b2e65c
NOTE, this is not the time or place to critique Chatbots (LLM); hold your thoughts.
Wide ranging discussions have considered aspects of the review. Whilst people quoted hallucination (humans hallucinate, machines have 'errors'), one conclusion is that although the review does-not fully or accurately represent the book, the content of the LLM review identifies valid, not fictitious, points of safety - probably drawn from other references.
As an alternative view, a 'human' web review was used for comparison, together with prompts exploring references. Again it was not possible to correlate this with the book, but from a safety viewpoint there is useful information to trigger individual thinking - a value of machine based searching.
https://chatgpt.com/share/689ae478-3...4-80bd34ab64e6
Before concluding too much;
https://chatgpt.com/share/687e3d50-9...8-b51a51b2e65c
NOTE, this is not the time or place to critique Chatbots (LLM); hold your thoughts.
Wide ranging discussions have considered aspects of the review. Whilst people quoted hallucination (humans hallucinate, machines have 'errors'), one conclusion is that although the review does-not fully or accurately represent the book, the content of the LLM review identifies valid, not fictitious, points of safety - probably drawn from other references.
As an alternative view, a 'human' web review was used for comparison, together with prompts exploring references. Again it was not possible to correlate this with the book, but from a safety viewpoint there is useful information to trigger individual thinking - a value of machine based searching.
https://chatgpt.com/share/689ae478-3...4-80bd34ab64e6
Before concluding too much;
- are the machine generated safety points valid and accurate wrt wide ranging human thoughts on these subjects ?
- does a LMM review help clarify - prioritise conflicting safety ideas promoted by humans ?
- how might this information alter opinion on the use of new technologies ?
- - is safety an illusion ?
Thread Starter
Joined: Jul 2003
Posts: 1,323
Likes: 54
From: An Island Province
The book
The book,'The Illusion of Safety'; free download
https://safetyconsultingroup.com/illusion/
“What feels safe isn’t always what is safe.”
https://safetyconsultingroup.com/illusion/
“What feels safe isn’t always what is safe.”
Joined: Mar 2006
Posts: 1,115
Likes: 86
From: England
What is safety
A CRM perspective
"… you get to decide how you consider, how to understand the situation.
It means being conscious and aware enough to choose what you pay attention to and to choose how you construct meaning from experience.
… isn’t about the capacity to think, but rather about the choice of what to think about."
https://www.dropbox.com/scl/fi/5ukgj...=h1rsldm5&dl=0
Academic definition - safety science
"No single precise definition of safety emerged from within the community, and thus a shared definition remains elusive."
Past - a condition or state. Present - activity which can be adjusted. Future - proactive, resilience
https://www.sciencedirect.com/scienc...25753525000232 (View pdf)
https://www.linkedin.com/pulse/one-w...aign=share_via
This (safety) is as good as it gets; don't disturb what we have achieved with new interventions, particularly if we don't know how we achieved 'safety'.
"… violations are symptoms of adaptation, not of loss of control."
https://www.dropbox.com/scl/fi/dnits...=ruovkdrs&dl=0
… and what ever it (safety) is called, lets change it - academics.
"Worry not. The early retirement does not refer to me. It is far too late for that. It refers to Safety-I and Safety-II which I earnestly want to retire and remove from public view. They are both by now old and tired concepts, that need some rest. Safety-I and Safety-II were thrust on an unsuspecting world in 2014. But ideas age faster than people, at least by one order of magnitude, so it is time to send Safety-I and Safety-II into retirement now, before they reach senility. They are replaced by decremental and incremental safety, which basically represent the same thing, namely two different approaches to achieve a state of safety (defined as having as few unexpected and unacceptable outcomes as possible). The newcomers will hopefully make it impossible to "infer" the existence of something like Safety-III or similar foolishness (although human folly so far seems to have no limits)."
Perhaps this is the background as to why we - the working industry - achieve little improvement; yet alternatively trying too hard might disturb what has been achieved.
An indication of a widening cultural gap in thinking about and between 'safety as imagined' by academics, and 'safety as done', by the workers.
https://incrementalsafety.com/oneweb...20oct%2007.pdf
https://incrementalsafety.com
Keep calm and carry on.
"… you get to decide how you consider, how to understand the situation.
It means being conscious and aware enough to choose what you pay attention to and to choose how you construct meaning from experience.
… isn’t about the capacity to think, but rather about the choice of what to think about."
https://www.dropbox.com/scl/fi/5ukgj...=h1rsldm5&dl=0
Academic definition - safety science
"No single precise definition of safety emerged from within the community, and thus a shared definition remains elusive."
Past - a condition or state. Present - activity which can be adjusted. Future - proactive, resilience
https://www.sciencedirect.com/scienc...25753525000232 (View pdf)
https://www.linkedin.com/pulse/one-w...aign=share_via
This (safety) is as good as it gets; don't disturb what we have achieved with new interventions, particularly if we don't know how we achieved 'safety'.
"… violations are symptoms of adaptation, not of loss of control."
https://www.dropbox.com/scl/fi/dnits...=ruovkdrs&dl=0
… and what ever it (safety) is called, lets change it - academics.
"Worry not. The early retirement does not refer to me. It is far too late for that. It refers to Safety-I and Safety-II which I earnestly want to retire and remove from public view. They are both by now old and tired concepts, that need some rest. Safety-I and Safety-II were thrust on an unsuspecting world in 2014. But ideas age faster than people, at least by one order of magnitude, so it is time to send Safety-I and Safety-II into retirement now, before they reach senility. They are replaced by decremental and incremental safety, which basically represent the same thing, namely two different approaches to achieve a state of safety (defined as having as few unexpected and unacceptable outcomes as possible). The newcomers will hopefully make it impossible to "infer" the existence of something like Safety-III or similar foolishness (although human folly so far seems to have no limits)."
Perhaps this is the background as to why we - the working industry - achieve little improvement; yet alternatively trying too hard might disturb what has been achieved.
An indication of a widening cultural gap in thinking about and between 'safety as imagined' by academics, and 'safety as done', by the workers.
https://incrementalsafety.com/oneweb...20oct%2007.pdf
https://incrementalsafety.com
Keep calm and carry on.
Last edited by PEI_3721; 19th February 2026 at 15:12.
Joined: Mar 2003
Posts: 187
Likes: 31
From: Brisbane
The Illusion of safety was a concept I came up with after years of studying aviation safety as a mere flight instructor.
In general terms, it means basically what that author says, which is briefly summarized by a Mark Twain quote - "It ain't what we don't know that gets us into trouble, its what we know for sure that just ain't so."
Perhaps a couple of simple examples will illustrate this point better. Since accidents still involve stalls and spins, there's good reason to suspect that pilots aren't trained to instinctively know when the wing is approaching stall, or how to keep the aircraft in balance without reference to the balance ball. Therefore, by teaching pilots how to recover from stalls or spins as we've always done, they're actually lulled into a false sense of security, thinking they know enough to recover, so no extra consideration or training is required.
Another simple example involves passenger screening at airports. The average pax may be impressed when their nail file or knitting needles are confiscated, thinking it's a sign the airlines are serious about their safety, or when flight attendants wake them up around TOPD to ensure their seat belts are fastened, tray tables stowed, and seatbacks upright, but there's really no safety benefit from doing any of that just after leaving cruise, or even before landing as long as ops are normal.
Too much is sold under the safety banner that doesn't seem much individually, but collectively distract valuable attention away from things that really matter.
In general terms, it means basically what that author says, which is briefly summarized by a Mark Twain quote - "It ain't what we don't know that gets us into trouble, its what we know for sure that just ain't so."
Perhaps a couple of simple examples will illustrate this point better. Since accidents still involve stalls and spins, there's good reason to suspect that pilots aren't trained to instinctively know when the wing is approaching stall, or how to keep the aircraft in balance without reference to the balance ball. Therefore, by teaching pilots how to recover from stalls or spins as we've always done, they're actually lulled into a false sense of security, thinking they know enough to recover, so no extra consideration or training is required.
Another simple example involves passenger screening at airports. The average pax may be impressed when their nail file or knitting needles are confiscated, thinking it's a sign the airlines are serious about their safety, or when flight attendants wake them up around TOPD to ensure their seat belts are fastened, tray tables stowed, and seatbacks upright, but there's really no safety benefit from doing any of that just after leaving cruise, or even before landing as long as ops are normal.
Too much is sold under the safety banner that doesn't seem much individually, but collectively distract valuable attention away from things that really matter.
