re: Is generalization killing creativity in the software industry? VIEW POST


This is wonderful!

I couldn't agree more, and I think you touch on several of the main antagonists of creativity in our modern world: a failing education model, and more generally our post-industrial capitalism.

But, zooming into code and generalization:
One of the wonderful things about specificity is its volume when carrying both explicit and implicit meaning. The right word, in the right place, is what distinguishes poetry from prose. Generalization is convenient, for reasons you mention, but generalization is also the evacuation of meaning and intention from the specific and tangible. More poetically, it is the departure from art and the arrival at utilitarianism.

It seems like if we can say anything about human intelligence, it is a strategic oscillation between the general and specific. We need to generalize and abstract concepts well enough to linguistically categorize them, while at the same time we need to map those abstractions onto the hard and dirty objects we bump up against in the world.

This reconciliation between the linguistically abstracted and the specifically encountered is absolutely what is missing in computer science. I do not think that it is a coincidence that AI research is hitting a hard wall at exactly this point (moving from rote recognition into meaning-making), and I would argue it's partially because computer scientists are bad at balancing between the two extremes of generalization and specificity. We certainly see the costs of that failing on other fronts, not least of all creativity - as you so well describe.

code of conduct - report abuse