vincenthsin

7 ms·

1. 1.▲
   by vincenthsin 1mo ago

   Even worse, it makes it difficult to distinguish between real and fake data.
2. 2.▲
   by vincenthsin 1mo ago

   As human-generated nutritional data runs dry, LLMs will end up eating their own garbage ?
3. 3.▲
   by vincenthsin 1mo ago

   Sorry,I mean capability limits.
4. 4.▲
   Ask HN: When do you think LLM capacity will reach its ceiling?

   4 points by vincenthsin 1mo ago|7 comments
5. 5.▲
   by vincenthsin 3mo ago

   LLMs should be used for translation and explanation instead of generating deterministic code. The source of software should be transformed from code to human requirements.
6. 6.▲
   by vincenthsin 4mo ago

   From the perspective of my own career in backend application development, the transformative impact that LLMs have brought to the field of application development over the past 1.5 years has exceeded that of the previous 10 (or even longer)