I work mostly with SMEs in the Midwest as a consultant, and I think people really miss just how shocking the overhead is on human resource related matters.
Caveat, I am not making a value or moral judgement, simply saying that from my perspective I can see where so much of this push for automation is coming from. Estimating conservatively, every employee is consuming approx. 1/2 their value (in terms of benefits) in time and management expenses, this includes turnover and solicitation for new replacement employees once they leave. I know businesses with 50 or fewer employees which are devoting outsized enough portions of their time in getting decent people you could make an argument they're in the temp business, not the manufacturing/service business.
Tyler Cowen at his blog has spoken about the 'productivity problem' we have globally numerous times and I really don't know what the solution is. Suffice to say it seems silicon valley is moving full-speed ahead with their solution, which is to simply replace the human.
Like Julian's comment above, there will be 1000 misses before the technology gets 1 big hit. But when it does, the snowball will be very hard to stop for industry once they really start crunching numbers.
Yes, if the landscape isn't smooth, the colonization process can be lumpier. Such as after passing over a mountain, and reaching a nice valley on the other side.
You mention the "slow steady process of automation", but I wonder whether one can also think if it in terms of the Schelling's Segregation Model.
In this approach a task is automated if and only if a critical mass of the surrounding task are also automated (i.e. there is a complementarity). The critical mass for automation can vary across different tasks depending on technological parameters etc.
The implication is that automation can be quite unpredictable and subject to big changes with large chunks of the task space becoming automated within a short period of time.
Another implication is that the creation of new tasks (or changes in the level of complementarity across task due to say regulation) can unleash unpredictable changes in the entire task space.
I work mostly with SMEs in the Midwest as a consultant, and I think people really miss just how shocking the overhead is on human resource related matters.
Caveat, I am not making a value or moral judgement, simply saying that from my perspective I can see where so much of this push for automation is coming from. Estimating conservatively, every employee is consuming approx. 1/2 their value (in terms of benefits) in time and management expenses, this includes turnover and solicitation for new replacement employees once they leave. I know businesses with 50 or fewer employees which are devoting outsized enough portions of their time in getting decent people you could make an argument they're in the temp business, not the manufacturing/service business.
Tyler Cowen at his blog has spoken about the 'productivity problem' we have globally numerous times and I really don't know what the solution is. Suffice to say it seems silicon valley is moving full-speed ahead with their solution, which is to simply replace the human.
Like Julian's comment above, there will be 1000 misses before the technology gets 1 big hit. But when it does, the snowball will be very hard to stop for industry once they really start crunching numbers.
Yes, if the landscape isn't smooth, the colonization process can be lumpier. Such as after passing over a mountain, and reaching a nice valley on the other side.
You mention the "slow steady process of automation", but I wonder whether one can also think if it in terms of the Schelling's Segregation Model.
In this approach a task is automated if and only if a critical mass of the surrounding task are also automated (i.e. there is a complementarity). The critical mass for automation can vary across different tasks depending on technological parameters etc.
The implication is that automation can be quite unpredictable and subject to big changes with large chunks of the task space becoming automated within a short period of time.
Another implication is that the creation of new tasks (or changes in the level of complementarity across task due to say regulation) can unleash unpredictable changes in the entire task space.