Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
est
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
1.
▲
by
est
5d ago
Here's how it started https://en.wikipedia.org/wiki/Murder_of_Poon_Hiu-wing A guy commited homicide and walks free after 19 months in jail.
2.
▲
by
est
11d ago
I used to search on google.ca exclusively with disabled js and cookie Until the day it no longer works. Sad.
3.
▲
by
est
18d ago
Stackless python had such potential, it could be extremely useful in Agent era for long running tasks with durable state. modern Python went in a somewhat different direction with asyncio, but with tasklet and continuation it could be a muc
4.
▲
by
est
19d ago
Very cool article. I had a similar idea where "fact checking" should be real programs for logic correctness. But IRL it's too vague. The exploit hunting is a better use case.
5.
▲
by
est
19d ago
Models are now having much tighter integration with harness. It's gunna to happen anyway. We are sailing into the age of vertically integrated intelligence.
6.
▲
by
est
26d ago
I find we are using AI wrong. AI shouldn't be "helping" with homework. AI should be the homework. Start with any topic or problem, students should be encouraged interact with AI, and learn stuff using Socratic method. Teacher
7.
▲
by
est
28d ago
I like the old MSN/GTalk "presence" idea Type your current status in a bar. And the thing that aggregates them? Twitter. People like to peek into what others are doing.
8.
▲
by
est
28d ago
I side with your opinion. CSS is bad for UI layout (maybe good enough for text-flows idk) ... opens the SO link yeah there's no answer to it (even if someone does, it would likely be closed by 5 others)
9.
▲
by
est
29d ago
seems like AI out of max_tokens
10.
▲
by
est
1mo ago
> but zero proof The repo gave you guide and benchmark of how to tweak ds4 pro for better performance, with a clickbait title (Fable 5 level), what kind of more proof do you require? More benchmarks? Reports for media outlets? STF model
11.
▲
by
est
1mo ago
This is cool What about the runtime size? I care this because I intend to run a stripped WASM version of DuckDB in browser.
12.
▲
by
est
1mo ago
I found the linked article "books don't work" from the quiz part highly interesting https://andymatuschak.org/books/ It explains a lot and works really well. I tried out in ChatGPT with a simple prompt:
13.
▲
by
est
1mo ago
I am thinking of AI credit resets, it forced me to finish my project because I already spent subscription so better not waste it! The cycle was a "process".
14.
▲
by
est
1mo ago
shameless self-plug: two hard things in computer science https://blog.est.im/2026/stderr-04 I think LLMs are reall good at conjecturing based on existing knowledge, but inventing new tools/lemma and new paradigm?
15.
▲
by
est
1mo ago
> incredibly simple WebSocket protocol Srsly bro? Websocket is simple? Talking about handling state persistency and connection partitions, such a headache.
16.
▲
by
est
1mo ago
The thing with LLM is they don't automatically pair parenesis/curly braces like we do with editors/IDEs. The closing } ) ] token has to be generated to match exactly the beginning { ( [ many lines before. You can challenge yo
17.
▲
by
est
1mo ago
Python has a less known advantage because it had no curly braces, so LLMs can focus its attention to logic instead of syntax. https://blog.est.im/2026/stdin-11
18.
▲
by
est
1mo ago
LLMs from years ago weren't aligned to death like these days. It's hard to get rid of "AI smell" than years ago.
19.
▲
by
est
1mo ago
Why do browser vendors have such a control over browsers? Can anyone make a dll injection and bring back manifest v2 ?
20.
▲
by
est
1mo ago
I made something similar Source: https://github.com/est/gitweets Demo: https://f.est.im/ The site and data is self-contained as a git repo. A git commit is a post. Blobs are pictures. Comments are stor
21.
▲
by
est
1mo ago
> What secret sauce do they have? Quant company usually squeezing every penny.
22.
▲
by
est
1mo ago
> even small errors compound over time and can lead to catastrophic outcomes So, death sentence even to frontier models?
23.
▲
by
est
1mo ago
Waiting for intelligence on a stick, plugin an USB, characters in, characters out. 100% local and no leaks.
24.
▲
by
est
1mo ago
> older stateful MCP (I’m going to call it “legacy MCP”) required two HTTP requests What? I never realized it was the case. I was calling my handcrafted MCP server without the header all the time!
25.
▲
by
est
2mo ago
Well I hoped it was some kind of SVG with alt text for copy. Turns out it's just a webp.
26.
▲
by
est
2mo ago
unfortunately JSON is doomed to fail in "the AI era" LLMs are surprisingly bad at generating JSON.
27.
▲
by
est
2mo ago
This is super cool. > Requirements: JDK 17+ and the Android SDK (platform 35). Bummer. I am stuck with Android 14 for the moment.
28.
▲
by
est
2mo ago
I feel LLM can only build what's known and popular. Sailing into the relms of lesser explored area are painful. It's like an intern but never learns, only getting replaced by better interns.
29.
▲
by
est
2mo ago
sounds like mathacademy and many other alike services
30.
▲
by
est
2mo ago
This is how reddit supposed to be. A universal comment box for del.icio.us
More ›