The Future of Knowledge Work
“Knowledge worker” is the term coined by Peter Drucker around the 1960's that describes the professional whose whole output is intellectual. While he or she does not contribute with muscle power and physical labor, the information they generate and the decisions they make help inform and potentialize the value an organization generates for society.
As the decades went by, it became clear that being a productive knowledge worker meant a mix between ability to think and ability to use tools for thought.
For example, in short time, electronic media all but replaced physical stores for information. Every knowledge professional today is reasonably expected to operate a computer. More critically, over time, professionals were expected to learn and take on the new tools of the day: be it the Office suite, web-based applications, and mobile applications, as required to do their work.
The shift that is most obvious today is the ability demonstrated by Generative AI tools to output knowledge work.
Depending on which side of the fence you're on, this means either: 1) completely replacing or massively displacing knowledge workers, as the same job will be done by the machine autonomously; or 2) just another tool to learn, just as any other software package.
Let me paint an even more nuanced opinion:
- The ability to automate drastically reduces the cost -- and thus the value -- of well-defined, repeating, and low-variability knowledge work.
- While the LLM is a language model, it lacks a model of the world itself. For the kind of tasks which require highly-specialized expertise built upon experience on the real world, the crop of AI technology displayed today is unlikely to make a dent.
- The current crop of AI technology also lacks any kind of reasoning, and can be wholly described as an imitation machine: it is extremely good at completing the most likely continuation. The intelligence we ascribe to it speaks more loudly of how we humans tend to project qualities upon it, and how we can let ourselves be “fooled” by plausibly and intelligently sounding arguments.