At its core: Thinking Knowledge Work

As human beings increase our output by inventing better machines, it worries us. Machine: a device that makes human work easy. How are we going to contribute value and get resources in return to live our lives, if this agent of ours, a machine, is going to do the work we have been doing for so long.

To think about this, you could visuialise current knowledge work as consisting of layers of abstraction.

We discovered we could light a fire from a spark. Lighting fire was work, a skill. For some, it was difficult. How do you apply force that rubbing two stones leads to a spark, and how do you use that spark to light fire on a combustible material.

Fast forward to gas stoves to ovens. We automated work and made it more efficient. But it also added more layers of technical expertise needed, more layers of abstraction. One or two people could light the fire.

But that was not the case with ovens and modern machines. There are various skills and several people whose knowledge, when combined, creates this machine. So, when it needs to be built or fixed, you need workers at the various layers of abstraction. Each worker possessed highly specialised knowledge and skillset - more specialised as you move down the layers. They memorised the parts that made the machine. They memorised the configurations those parts could be brought together in, and configurations that they should not.

Then they assesmbled the parts.

We can drill all the way down to individual parts of the oven, and the subparts and so on. Each iteration follows the same model - the lower the layer of abstraction, the more specialised the knowledge and skillset needed.

A massive part of this work was:

1. Memorisation. Memorisation of the parts that made the machine. Memorisation of valid configurations. An individual brain can only store so much. An individual has only got so much time.

Remember the equation: work = force x distance, or Power x Time. For an individual, power and time are limited and capped.

2. The other part was execution, putting the memorised parts and configurations into use.

Could it be argued that as we invented more and more and understood the value of combining knowledge to break the limits of invention on an individual, we went down the complex layers of abstraction? This limitation was due to time, resources, knowledge, mental capability, and limited experiences of an individual as opposed to groups of people with different experiences

A vast majority of human workforce got allocated to lower layers of abstraction. They developed specialised knowledge. Seeing the big picture was just not what they did. So, only a fraction of humans were now thinking of lighting fire to solve the cold and hard-to-digest food problem.

Most were focused on the sub-problems, like the optimal amount of force needed, the most combustible material to turn the spark into a fire, etc.

Most of our focus went to optimising the current solutions. That's where the effort and time went. That was the work we produced. We optimised problems and became really good at it.

But was it at the cost of pushing barriers? What would happen when the vast majority focusing on optimising solutions could focus on solving more problems? What happens then when we build a machine - AI extensions of knowledge workers - that can take care of the majority of the lower technical layers?

Do we run out of work?

In the bigger picture, the human resource that previously focused on optimising existing solutions has to spend time somewhere.

What happens to that workforce?

Could they start to move up, be a little more abstract, and think about more solutions rather than optimising the existing ones now that their AI extension can do that?

Are we going to see problems we did not realise existed (or did not imagine as solvable) being solved?