CT,
new tools (AI) encourage people to look for a problem where the tool can be applied, irrespective of relevance to the end objective; safety.
Focus on human conversation assumes that the human is a problem; and because there is a problem the relevant contributing factors can be identified. Extensive research so far has yet to find a satisfactory understanding of human behaviour (for safety purposes); adding more ‘big data’ will not necessarily identify anything better, although it could provide an alternative view.
The critical issue is who looks at the alternative view; if this involves the original assumption then those people will find what they were looking for - human error, due to their own human bias.
It may be more beneficial for aviation safety to record the assumptions made at management and regulatory levels and asses these conversations for patterns, and if these match what actually happens in operation - how people at the sharp end create safety, vs those who think that they regulate safety.
megan,
but managers may have the skill to let hot bricks cool, the use them to build a higher defensive wall.
D479,
nothing to fear except fear itself.