Peas in a Pod

I was actually as struck by the similarities between Kurzweil and Fukuoka as the differences.

In particular, many of the algorithmic techniques that Kurzweil seemed most optimistic about were emergent in some way – algorithms that learned based on many iterations of “experience”.

In some sense, isn’t the ideal farmer also an example of such a learning mechanism? That can appreciate the delicate and subtle interactions of many insects, weeds and natural species?

Or, is there something missing? Humility perhaps?

Im also reading Jane Jacobs at the moment, and I was struck by some of the overlaps between her thinking and that of Fukuoka – of respecting “natural” diversity, of “optimal” “solutions” arising out many individual micro-interactions, and of the best heuristics as those being learned over many iterations of trial and error.

Is that really so different from what Kurzweil is proposing? Or, does the very idea of databases, computation and the abstraction that it entails reduce and simplify in a way that loses the essence?

2 thoughts on “Peas in a Pod

  1. Doesn’t this assume that the knowledge of the farmer can be entirely captured by machine implementable heuristics? Fukuoka seems interested in contextualizing knowledge, and identifying those things which cannot be easily codified or quantified. It’s maybe a bit much to call this intuition, but he’s definitely interested in that which cannot be teased out through empirical approaches.

  2. Agreed. The context seems vital. That’s not to say that a million iterations on a given problem won’t yield a kind of context — but perhaps not the broad one the farmer experiences over a lifetime. Or maybe it could, if informed by the right parameters that encapsulate the farmer’s memory and physical interactions with the land. Seems far off, hard to imagine, but within Kurzweil’s vision.

Comments are closed.