3 ms·
What bothers me is this: Popper once argued that to prove something, you need to falsify the claim. I am reminded of the story about "no such thing as a black
by gardenfelder 4y ago
What bothers me is this: Popper once argued that to prove something, you need to falsify the claim. I am reminded of the story about "no such thing as a black swan", which held right up until someone brought a black swan into town from afar.
- karmakaze 4y agoInstead of extrapolating "there are no black swans in town" to "the world", it would be better to calculate the probability of no black swans being in town (given that there are N swans in town), based on hypothetical populations of swans and varying uniformity of distribution. A zero-knowledge proof[0] uses similar reasoning. [0] https://en.wikipedia.org/wiki/Zero-knowledge_proof https://en.wikipedia.org/wiki/Zero-knowledge_proof
- eesmith 4y agoWhen this came up yesterday, at https://news.ycombinator.com/item?id=34130864 https://news.ycombinator.com/item?id=34130864 , I pointed to an article by Nancy C. Knight, "No Two Alike?" in Bulletin of the American Meteorological Society, 1988. https://journals.ametsoc.org/downloadpdf/journals/bams/69/5/1520-0477-69_5_496.pdf https://journals.ametsoc.org/downloadpdf/journals/bams/69/5/... . > Figure 1, a photograph of actual snow crystals, not replicas, illustrates a striking example of two snow crystals which, if not identical, are certainly very much alike. ... The two very similar snowflakes are in the bottom center of figure 1, side-by-side.