ChatGPT Is a Blurry JPEG of the Web
Chiang's New Yorker essay arguing a large language model is a lossy compression of its training corpus. Management Craft cites it for the consequence that a model's picture of 'great' is the compressed market average, never your company's localized standard.
- Author
- Ted Chiang
- Published
- 2023
- Original
- Read it at newyorker.com