I find the comparisons between AI and humans laughable. It’s an apples and oranges delusion.

It is nonsense to even try to compare the embodied fallible human with a dis-embodied mechanical computer/machine. Even when it comes to language acquisition and Large Language Models (LLMs) most who write about it have no expertise in language/linguistics acquisition. Indeed, most hold to a brain-centric understanding of intelligence, which is why we call it ‘artificial’.

If you think humans are a computer on top of a body then be fearful that AI will take your job. However, if you know humans as embodied fallible beings who live in social e-motional relationships, AI will never take your job.

If you think that safety is about the repetition of paperwork, then you will think AI can take your job. If however, you know that safety is about embodied listening, relationships, social engagement, connecting and thinking critically, AI will never take your job.

If you have no expertise in culture, anthropology, social psychology or ontology, then you might fear that AI can take your job. If however, you understand life and being as a shared world (Durt, Fuchs and Tewes (2017) Embodiment, Enaction, and Culture, Investigating the Constitution of the Shared World) where thinking requires embodied intelligence (Claxton, Robinson, Fuchs etc) then AI will never take your job.

This is why the metaphor of brain-as-computer needs to be rejected by Safety (https://safetyrisk.net/safety-is-not-all-in-the-brain/). This is why the One Brain Three Minds model in SPoR is critical for understanding risk (https://safetyrisk.net/one-brain-three-minds-what-the-research-says/).

If you view safety as a mechanistic process then you might fear AI will take your job. If however, you understand safety as a relational and experiential job, AI will never take your job.

This is why understanding an ethic of personhood is critical for understanding safety (https://www.humandymensions.com/product/the-ethics-of-risk/). This is why an engineering focus on risk will never understand the nature of culture, risk and being. Yet in safety we have this strange idea that the best to deliver on ethics is an engineer???

This is also why so much of the hype and propaganda about AI in safety is about faith in AI not an understanding of the consciousness of e-motional, fallible personhood. It is also why an understanding of risk and safety as a wicked problem, takes any hope expressed in AI as myth.

The key to all this is what you understand the safety job to be? And is the safety industry preparing people for such a job? If you look at the safety curriculum and all of the training for post-nominals, the answer is no. For more on AI and safety, read here: https://safetyrisk.net/faith-in-ai-and-ai-safety/