Automation causes risk to supply of experts

[sjurl]

Aspect	Description
Challenges	There is a perceived risk that automation including some machine learning processes may automate some processes to a stage where training new experts will become difficult. This will in turn make it difficult to provide good classification information for training in machine learning.
Action required	​​DIG to Monitor. Discuss with WGMLEARN on their thoughts on communicating development of machine learning, to consider the issue that there could be worries.
Ownership	No specific owner

legacy entry

[jensr]

Low likelihood since volume of data generally collected is increasing, and ML functions are predominantly used to meet the gap in analytical demand. Generally most ML support tools still require extensive validation and training from human expert

The impact though would be very high if there was a loss of experts in a field. New developments of capabilities would be threatened.

[OliverWilliamsDataManager]

Check if there is an existing appropriate opportunity, if not create one. Link back to this one.