Developer publishes his product's underlying dataset as open data instead of keeping it secret
A developer who built a wedding speech generator decided to release the dataset behind its checks, a list of Japanese taboo words, under a CC-BY-4.0 license while keeping the prompt private. He argues that compiled public cultural knowledge cannot function as a real moat, and that hiding the data made his code harder to test and hid a latent bug. Making the list open, he says, also seemed to bring no measurable downside, since few people would build the whole product from a small JSON file. He separates facts, which he publishes, from product judgement about tone and structure, which stays private. He concludes that the real test of a moat is whether a competitor could reproduce it over a weekend.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)
Log in to join the discussion and vote.
Log in