hubraumhugo

10 ms·

31.
by hubraumhugo 5mo ago
Appreciate the feedback, just updated the title to be more clear.
32.
Scoring Show HN submissions for AI design patterns(adriankrebs.ch)
333 points by hubraumhugo 5mo ago|235 comments
33.
Scoring 500 Show HN pages for AI design slop(adriankrebs.ch)
4 points by hubraumhugo 5mo ago|1 comments
34.
Show HN: Quant Job Market Visualizer(quant.kadoa.com)
1 points by hubraumhugo 5mo ago|0 comments
35.
Apply Less, but Better(adriankrebs.ch)
3 points by hubraumhugo 6mo ago|1 comments
36.
The dead Internet is not a theory anymore(adriankrebs.ch)
419 points by hubraumhugo 7mo ago|307 comments
37.
Therac-25(en.wikipedia.org)
3 points by hubraumhugo 7mo ago|0 comments
38.
Vercel deployments were created without Middleware(vercel-status.com)
1 points by hubraumhugo 7mo ago|1 comments
39.
by hubraumhugo 7mo ago
In a separate email, they note: > For affected deployments created during this window, any requests served until 21:10 UTC did not execute Middleware. The issue has been fully remediated and no redeployment is required. > If, for auth
40.
Ask HN: How is your team changing documentation for AI-assisted coding?
3 points by hubraumhugo 7mo ago|1 comments
41.
by hubraumhugo 7mo ago
It's working pretty well for generating an xkcd comic for your HN profile: https://hn-wrapped.kadoa.com/ Previous nano banana frequently made speech attribution errors, the new one seems a lot more consistent.
42.
AI coding broke the Ballmer Peak(adriankrebs.ch)
1 points by hubraumhugo 7mo ago|1 comments
43.
The New Ballmer Peak(adriankrebs.ch)
3 points by hubraumhugo 7mo ago|1 comments
44.
by hubraumhugo 8mo ago
Anybody else not seeing it available in Codex app or CLI yet (with Plus)?
45.
by hubraumhugo 8mo ago
Kadoa | Multiple roles | Remote | Full-Time | https://kadoa.com Web scraping hasn't changed in decades: engineers still write brittle scripts that constantly break. Kadoa automates the entire web data pipeline with AI agent
46.
by hubraumhugo 8mo ago
I recently met a European space startup founder and was surprised to learn how much space innovation is happening in Europe with ESA. Europe wants to become less depended on SpaceX and NASA, and is heavily investing there. More funding + st
47.
By almost every measure, life is better than ever.(adriankrebs.ch)
2 points by hubraumhugo 8mo ago|8 comments
48.
Human Progress Data(humanprogress.org)
2 points by hubraumhugo 8mo ago|0 comments
49.
Ask HN: What was your best interview experience?
2 points by hubraumhugo 9mo ago|1 comments
50.
by hubraumhugo 9mo ago
dang's explanation sums it up nicely: > It's human nature: https://en.wikipedia.org/wiki/Negativity_bias . Everyone does it, but we perceive other people as doing it more than we do, which is itself a varia
51.
by hubraumhugo 9mo ago
Politics aside: If this is all true and was a snatch and grab, it will go down as one of the most impressive military operations in the 21st century.
52.
by hubraumhugo 9mo ago
Kadoa | Multiple roles | Remote | Full-Time | https://kadoa.com We’re on a mission to give humans and LLMs reliable and fast access to web data. Web scraping used to be the same for decades (brittle scripts that break constantly
53.
A16Z big ideas 2026: Part 1(a16z.com)
2 points by hubraumhugo 9mo ago|0 comments
54.
by hubraumhugo 9mo ago
Merry Christmas HN! You're one of the few constants among the many variables in my life, please never change :)
55.
Microsoft wants to replace its C and C++ codebase, perhaps by 2030(theregister.com)
1 points by hubraumhugo 9mo ago|4 comments
56.
by hubraumhugo 9mo ago
"The Thinking Game" is an absolutely fascinating and inspirational documentary about DeepMind and Demis Hassabis: https://www.youtube.com/watch?v=d95J8yzvjbQ Makes you really optimistic about the future of humanit
57.
by hubraumhugo 9mo ago
Would love to learn more about how this is built. I remember a similar project from 4 years ago[0] that used a classic BERT model for NER on HN comments. I assume this one uses a few-shot LLM approach instead, which is slower and more expen
58.
by hubraumhugo 9mo ago
Thanks. I now run a two-step process: first pass reads through all posts and comments to extract patterns, second pass uses those to generate the content. Should be much more representative of your full year now :)
59.
by hubraumhugo 9mo ago
Ah, I've found the issue. Turns out I didn't account for case-insensitive HN usernames like yours :) should be fixed, love your current xkcd :D
60.
by hubraumhugo 9mo ago
Appreciate the feedback, will try to iterate it to greatness further. It's still a bit hit or miss, but I've made a few improvements: - improved prompts with your feedback - added post/comment shuffling to remove recency bias

More ›