News for hackers. Stuff that matters.

ML0037

This article just explained the behaviour I’ve been seeing in my grandfather. He has dementia and used to hunt wild boar (very common in central Italy), but he knew the local wildlife well. Since developing dementia, for obvious reasons, he unfortunately hasn’t been the same. But a group of doves has nested on the gazebo on his terrace, and every year they come back (either them or their offspring). He watches them all day, giving them food and water and following every stage of their growth. It’s beautiful to see nature indirectly helping humans who, let’s be honest, at least in his case, hadn’t done much good for nature.

Ask HN: Can LeetCode (or similar) keep critical thinking sharp in the AI era?

These days, at least in my current situation, LC feels less like interview preparation and more like a mental gym for technical thinking, i don't like the LC grind, so i just do 1/2 problems per week just for fun. Have you felt this too, or are there better ways to develop (don't lose) this way of thinking?

( Read More... | 2 comments | 2 points | on HN )
Hello,

Just for your information, when I open "https://claude.ai/new" the script automatically open a random article, do you know why?

Re:Om Malik has died on Friday June 26, 2026 @11:48AM (#48688048)
This was his last post: https://om.co/2026/06/08/taking-a-few-days-off/ He seemed to know that something wasn’t right. I really loved his writing.

Rest in peace.

Show HN: I built a small audit layer for LLM-as-judge decisions

https://github.com/MatteoLeonesi/claim-memory-graph-sdk

I made this while checking model graded answer and helped me to check the odd cases by hand. Not sure if it’s useful to anyone else.TL;DR: it breaks an LLM judge run into claims->evidence->verdicts and flags when a verdict is not supported by the evidence, so i can check it manually.

( Read More... | no comments | 2 points | on HN )

Show HN: I made a small helper for checking model-graded answers

https://github.com/MatteoLeonesi/claim-memory-graph-sdk

I made this while checking model graded answer and helped me to check the odd cases by hand. Not sure if it’s useful to anyone else.TL;DR: it breaks an LLM judge run into claims->evidence->verdicts and flags when a verdict is not supported by the evidence, so i can check it manually

( Read More... | no comments | on HN )

Show HN: I made a small helper for checking model-graded answers

https://github.com/MatteoLeonesi/claim-memory-graph-sdk

I made this while checking model graded answer and helped me to check the odd cases by hand. Not sure if it’s useful to anyone else.TL;DR: it breaks an LLM judge run into claims->evidence->verdicts and flags when a verdict is not supported by the evidence, so i can check it manually.

( Read More... | no comments | 2 points | on HN )
Hi! for ChatGPT I go in settings -> Data Controls -> delete all chat

Show HN: Script to bulk delete Claude chats from the web UI

https://github.com/MatteoLeonesi/bulk-delete-claude-chat

I haven't found a way to delete all chats in bulk like you can on Chatgpt. With Claude, you have to scroll to the bottom, select everything, and delete. The problem is, if you have a lot of chats, it becomes impossible. I created this script. It does it alone. I hope it helps someone.(conversations disappear from the UI slowly, over several minutes, and remember to keep the tab open until the...

( Read More... | 22 comments | 19 points | on HN )
i still code, but i'm searching an IDE that preserve the human in loop, for example showing better the code added/changed by the agent, but yeah i would like to have a better interaction with the agent while coding, infact i tend to use small models with high inference so i can procede with him without losing the flow.
Re:The Git Commands I Run Before Reading Any Code on Wednesday April 08, 2026 @01:47PM (#47693717)
i’ll try to use the in an hook and test them with Claude. Thank you !

Show HN: Yumi Reader – Turn any article into distraction-free pure text

https://github.com/uscne/Yumi-Reader

This browser extension helps by turning articles into text-only reading views using Mozilla's Readability library with custom CSS informed by accessibility and readability research. It follows four design choices based on W3C WCAG 2.2 standards: a sepia background to reduce eye strain, 1.5× line spacing for better readability, 50–75 character line length to reduce eye fatigue, and sans-serif...

( Read More... | no comments | on HN )