Work in the Time of AI

About 38% of U.S., 35% of German, 30% of U.K., and 21% of Japanese workers could become unemployed due to AI and robotics. Widespread job loss from the application of technology seems inevitable. Automation is not designed to replace people but to conduct a process more efficiently. That business leaders choose to apply automation to increase productivity through the exchange of human labor with machine activity is a different matter. The trends are based on human choice and not so much the opportunity machines make available.

Machine Intelligence @ GE

Fareed Zakaria GPS spoke to Jeff Immelt who sees the need for improvements in manufacturing. The machines GE builds need to get smarter, have more intelligence, and be wired up in such a way that failures can be predicted and performance improved. This is expected to grow the value of the devices, services, and cargo craft, turbines and other machines GE sells.

I remember GE telegraphing these themes for the last 10 years. At the time I remember thinking that they may be on to something. Now, I am less sure. Ten years is a long time in human accounting of progress. Maybe some things are hard because Jeff explained to Fareed that GE has to get better at doing the industrial machine digital machine integration.

The subject turned to an idea I’ve advocated for the last 3 to 4 years. I call it the Red Hat Business Model, but whatever you call it is the future way of doing business. Jeff or Fareed stated it as “Selling big data and service after you sell the jet engine”. Swap jet engine for anything and that is the quintessential business model of the coming decades. I will expand on that and say it is no longer about who makes what but about who provides the best “interpretation” of information.

Humans want to be smarter. Greater intelligence provides a survival advantage. That is hardwired into human DNA. Profitability derives from improving the efforts of others and the most superior way to do that is not through conveyor belts, trucks, software, industrial or digital machines per se. Those are just gateways to information, to process, output you study in order to evaluate where you are and where you need to be to remain productive.  Therefore, curating information and having useful information to offer is the business …

Anyway, Jeff sees several phases in the transformation of industrial commerce. The 1st phase is the concept of the smarter worker. A person as good with a computer as they are with a welding torch. Other phases he cannot identify yet. Despite a good prescription for what GE needs, I do not think the transformation is going to work. I actually advocate the expansion of domain experts in the application of IT but I don’t see that model scaling to accommodate a large enough workforce.

Other points he made:

  • Tech boosts factory output
  • Good tech practice helps you get it right the first time
  • More productivity raises wages
  • Today, displaced workers tend to make less in a new job. This creates anger

Jeff concluded with quality exports to the rest of the world is vital to growth. The goal is to make better things so people across the sea trade with you. I wonder if digital machines will neutralize gains as history shows all advantages equalize out.

Do Machines Take Over?

Human laziness will actually prevent machines from displacing everybody. Humans are inclined towards the status quo and thus many businesses will be years from now as they are today. True enough, you do not need 100% machine expansion to result in severe employment problems for society. If even 30% to 50% of people cannot find work is enough to make their lives more difficult.

A major wildcard is Amazon the large web-based retailer. A large segment of civilization buys products and solutions from them. Some of their factories are already robotically operated. Drone delivery is something they are pioneering. If Amazon succeed in these efforts to allow them to more easily overtake other retailers, then they create incentive for many businesses to keep up by doing similar things.

You do not need as many digital laborers as physical so while there will be job opportunities, they will be a fraction of the previous requirements for physical labor. More people will be out of work. Amazon’s foray into automated delivery and fulfillment may launch a chain reaction. Organizations who presently see no need for this level or kind of automation may see it as a matter of survival. An automate everything mindset may grow resulting in more people failing to find work.

What is the New Work?

Alternatives do exist. In a few weeks or so I will be writing about a more detailed approach. For now, the easiest ideas to consider are those experiments underway in different parts of the world involving UBI. The proper implementation of UBI may be just the right measure to provide a clean transition to the future. An ideological shift as large as the onset of the agricultural age which proved massively beneficial to humans overall.

Meanwhile, economic focus can shift away from industrial output (as the machines takes care of that) to “soft sectors”. What we call leisure, education, personal services, boutique products, certain kinds of “interpreted information” services, and the rough edges around automation gain emphasis. Industries like these may or may not be mass consumption industries, but with the right classification, identification, and educational preparation of more people in combination with UBI, there could yet be work for everybody. More extensive design may be required, but this could be a good patch on Human Civilization 2.0.

Related Conversations

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s