Comment by PaulHoule
2 years ago
I thought the author was uncharacteristically perceptive for a reporter. Yann LeCun or Geoff Hinton couldn't have come up with a better analogy.
2 years ago
I thought the author was uncharacteristically perceptive for a reporter. Yann LeCun or Geoff Hinton couldn't have come up with a better analogy.
The author is not a random reporter but Ted Chiang, a well-known science fiction author. The movie "Arrival" is based on a story by him.
Which explains why this is being promoted:
He paid for an advertisement, wrote this article as that advertisement or had it ghostwritten, and now it's being hyped.
Analogy of?
"Blurry JPEG" for how ChatGPT "compresses" character-based knowledge into vectors. That "compression" process gives ChatGPT an ability to generalize because it learns statistics (unlike JPEG) but like JPEG it is a lossy process.
It's a terrible analogy because the entire point of ML systems is to generalize well to new data, not to reproduce the original data as accurate as possible with a space/time tradeoff.
10 replies →
So they can't afford an actual subject-matter expert for their articles?
In a world where supposedly more-tech-industry-aware writers are talking about what "ChatGPT believes" and other such personification... show me a better article.
Any article on how ChatGPT works would be much better.