"...While I guess, since engineering is essentially a discipline of decision making, there will probably still be vocations where their skills are important and there will always be crooked politician slots available for former lawyers, and bone-sawing will be around for the doctors, accountants may find themselves in a pile of slide rules..."

I believe the changes will be MUCH more significant than that implied above.

Off the top of my head:

President and executive officers review and present concepts 25% of the year as a status update, repeating information previously presented roughly 6 times
VP level employees make decisions roughly 40% of their time related to aggregating, refining and decision support for what should be entirely automated processes.
VP, Director and manager level employees spend 16 out of every 22 working days every month on period open, inventory and period close activities. They request support from numerous individuals along the way.
VP, Director and Manager level employees spend 30% over 4 months to iterate and review and align on forecasts and plans that should be 80% automated.
Supervisors, managers and location P&L leaders spend 25% of everyday recounting and summarizing information for decision support.
Trainers, Subject Matter experts and senior hourly staff spend 8 weeks on every new hire to accumulate knowledge and bearing on safety, rate and quality.
Senior technical and Operations leaders and subject matter experts spend 70% of their time realigning and troubleshooting processes that fall far from benchmark (prior proven performance).
VP, Director and location managers iterate daily on Supply chain support and visibility related data analysis and decision making when data for sales and operations planning is accurate and timely.
Engineers, drafts people and technicians revalidate, rerate and confirm relatively simple calculations with data that is in the system already.
Operators on mostly automatic machinery spend most of their time watching to ensure no problem occurs. With a few algorithms looking after the system, their number and their required experience will be reduced significantly.

and on and on and on.....

With no humanoid robots running around, there will be very few jobs required when AI is fully empowered:

For a location:

Maintenance
Non-automated material handling,
tooling set up,
loading and unloading,
security
Location manager (now combined with maintenance, operations)
Location 2nd (now combined with packing, material receipt and warehouse)

For the "home office"

80% less VPs
70% less Directors
70% less managers
30% MORE Product/special Project Managers (reassigned from other roles)
200% MORE Data Scientists (borrowed from MI SME and industry)
500% MORE Data Engineers (borrowed from Tech corps)
90% less FP&A staff

This is a human generated analysis, so, it will likely be replaced.