missmoss

4 ms·

1.
by missmoss 2mo ago
This is real. I was very angry that AI kept breaking the rules I wrote, so I asked Claude to crawl its own history logs. And then I found: every time after it breaks a rule, the chance of breaking rules goes up. I feel this is like few-shot
2.
by missmoss 2mo ago
My guess is AI mostly increases the throughput of low-quality output, and the cleanup burden lands on the people with better judgment
3.
by missmoss 2mo ago
I love this already. Slow-moving large animals are an instant yes for me
4.
by missmoss 2mo ago
This makes perfect sense. My friend's cat is basically into window TV too
5.
by missmoss 2mo ago
Love that sound. I could sit at the beach all day
6.
by missmoss 2mo ago
Weirdly soothing to watch it slowly reposition itself
7.
Ask HN: What livestream do you keep open in a tab?
58 points by missmoss 2mo ago|21 comments
8.
by missmoss 2mo ago
This is neat! It tackles a boring but real problem with agents. When you have too many tools, the model gets confused. This is a good search box for its tools instead of dumping everything into the prompt. Benchmarks is amazing!
9.
Notes from My First Product DS Mock Interview(clairetsao.substack.com)
1 points by missmoss 2mo ago|0 comments
10.
Loop Engineering: Design the Loop, or Be the Loop(clairetsao.substack.com)
2 points by missmoss 2mo ago|0 comments
11.
For the First Time, AI Was My User(clairetsao.substack.com)
2 points by missmoss 4mo ago|0 comments
12.
Mermaid Diagrams Are Unreadable in Real-World Technical Docs(clairetsao.substack.com)
2 points by missmoss 4mo ago|0 comments
13.
by missmoss 4mo ago
Cool! I watched the demo. The concept makes sense. Quick thought: free-text answers might be high friction for adoption. Not sure if you'd consider various verification method, but something like multiple choice questions or "high
14.
by missmoss 4mo ago
When working on an existing project, I always take a close look on what got changed by LLM. Sometimes I explicitly tell it which line to change and nothing else. Like what most people found, LLM sometimes, if not always, creates redundant o
15.
by missmoss 4mo ago
Curious if there's a demo/screenshot on what questions developers should expect and what acceptable answers look like. Checked out the repo but didn't find the showcase.