dawndrain

3 ms·

1. 1.▲
   by dawndrain 2mo ago

   Epoch AI estimates that LLM costs are falling by over 10x per year at fixed intelligence https://epoch.ai/data-insights/llm-inference-price-trends , so I expect prices to fall at this level of quality. I also think ther
2. 2.▲
   by dawndrain 3mo ago

   In addition to the repo I have a long blog post reflecting on my experience here, including war stories, more detailed costs, and bloopers https://dawndrain.notion.site/Making-AI-Movies-39cbb81dad798...
3. 3.▲
   Show HN: A Pipeline for Making 10-minute AI Movies with Claude Code and Seedance(github.com)

   21 points by dawndrain 3mo ago|4 comments
4. 4.▲
   by dawndrain 5y ago

   Certainly lots of things to learn still. Going point by point: It's not "he" it's "she". I don't think it's controversial that you can be much more concise with python. My experience first learning Ja
5. 5.▲
   by dawndrain 5y ago

   > I was hoping for more DS related stuff. It almost sounds like you're learning to be a SWE! It kind of felt that way too :). Some more data sciency things were learning how transformers work, hacking pytorch, using visualization to
6. 6.▲
   by dawndrain 5y ago

   > did you do any DD other than taking his advice? I did some backtesting simulations that made leveraged investing look pretty awesome. The effective borrow rate for funds like spxl is crazy low, way better than if I were to borrow mysel
7. 7.▲
   by dawndrain 5y ago

   I heard this second-hand, not totally sure it's true