Hacker News

Article URL: https://thereader.mitpress.mit.edu/bitching-betty-speaks-how-talking-machines-got-their-gendered-voices/

Comments URL: https://news.ycombinator.com/item?id=43588315

Points: 1

# Comments: 0

Article URL: https://scitechdaily.com/scientists-just-discovered-quantum-signals-inside-life-itself/

Comments URL: https://news.ycombinator.com/item?id=43588301

Points: 1

# Comments: 1

Article URL: https://www.nature.com/articles/s41587-025-02635-7

Comments URL: https://news.ycombinator.com/item?id=43588288

Points: 1

# Comments: 0

Article URL: https://lazyreminder.com/

Comments URL: https://news.ycombinator.com/item?id=43588253

Points: 1

# Comments: 0

Article URL: https://mitp-content-server.mit.edu/books/content/sectbyfn/books_pres_0/6515/sicp.zip/full-text/book/book.html

Comments URL: https://news.ycombinator.com/item?id=43588218

Points: 1

# Comments: 0

Article URL: https://www.psychologytoday.com/us/blog/living-single/202504/dementia-is-more-common-among-the-married-than-the-unmarried

Comments URL: https://news.ycombinator.com/item?id=43588208

Points: 2

# Comments: 0

Article URL: https://www.bps.org.uk/research-digest/how-make-habit-stick

Comments URL: https://news.ycombinator.com/item?id=43588181

Points: 2

# Comments: 0

Article URL: https://en.wikipedia.org/wiki/Carom3D

Comments URL: https://news.ycombinator.com/item?id=43588174

Points: 1

# Comments: 0

Article URL: https://www.scattered-thoughts.net/writing/a-rough-survey-of-compilation-recompilation-and-compile-time-evaluation/

Comments URL: https://news.ycombinator.com/item?id=43588167

Points: 2

# Comments: 0

Article URL: https://www.bps.org.uk/research-digest/do-women-really-talk-more-men

Comments URL: https://news.ycombinator.com/item?id=43588157

Points: 1

# Comments: 0

Article URL: https://www.nytimes.com/2025/04/04/opinion/elon-musk-doge-technocracy.html

Comments URL: https://news.ycombinator.com/item?id=43587789

Points: 2

# Comments: 0

Hi HN! I've built SwiftHive, a private Swift package registry that solves several pain points with Swift's dependency management. We drastically reduce build times, ensure reproducibility, and I'm actively looking for beta testers from the community.

Comments URL: https://news.ycombinator.com/item?id=43587780

Points: 1

# Comments: 0

Article URL: https://github.blog/changelog/2025-04-04-github-mcp-server-public-preview/

Comments URL: https://news.ycombinator.com/item?id=43587771

Points: 2

# Comments: 0

Article URL: https://www.openculture.com/2025/04/watch-bob-dylan-make-his-debut-at-the-newport-folk-festival-in-colorized-1963-footage.html

Comments URL: https://news.ycombinator.com/item?id=43587766

Points: 1

# Comments: 0

I'm living on a small SaaS that will hopefully sustain myself into retirement and give a little income (have had some past exits that I am lucky for, this one will probably cap out on TAM, but good enough)

That said, I'm funding/developing this all myself and I'm looking to move myself and the company out of the US.

I'm sure the US is still the top place to run a startup for financial/networking reasons, but I don't need that for this one and I'd rather deal with a little more red tape in a country I feel aligns better w my values.

I'm just curious if anyone has made this kind of move , and any advice on how to do it right for a small 1-2 person profitable LLC?

Comments URL: https://news.ycombinator.com/item?id=43587764

Points: 2

# Comments: 0

I heard reading well-written code can improve my skills. Instead of reading top-rated projects on GitHub, which I do, I thought to ask the community to share their favorite OSS they believe to have examplary code.

Comments URL: https://news.ycombinator.com/item?id=43587758

Points: 1

# Comments: 0

Article URL: https://www.youtube.com/watch?v=TX65XCk2-6U

Comments URL: https://news.ycombinator.com/item?id=43587742

Points: 1

# Comments: 0

Article URL: https://docs.fcc.gov/public/attachments/FCC-25-22A1.txt

Comments URL: https://news.ycombinator.com/item?id=43587737

Points: 1

# Comments: 0

Article URL: https://ai.redpanda.com/

Comments URL: https://news.ycombinator.com/item?id=43587736

Points: 1

# Comments: 0

I built Kotlin-bench, the first benchmark for evaluating LLM performance on real-world Kotlin and Android tasks.

Current AI benchmarks like SWE-bench primarily focus on languages like Python or JavaScript, leaving Kotlin and Android development largely overlooked. Kotlin-bench addresses this gap by providing a clear metric for evaluating AI models specifically on Kotlin.

Like SWE-bench, Kotlin-bench assesses AI models using real GitHub issues from widely-used Kotlin repositories such as kotlinx.coroutines and Anki-Android. Solutions are objectively validated by running each project's unit tests—a model passes only if its generated code correctly resolves the original issue.

The dataset includes 100 GitHub issues selected from well-maintained and thoroughly tested repositories.

Major takeaways:

- Gemini 2.5 performed best, solving 14% of tasks, closely followed by Claude 3.7 (thinking mode).

- "Thinking" models substantially outperformed standard models, with Gemini 2.5, Claude 3.7 thinking, and Deepseek R1 topping our charts.

- Surprisingly, OpenAI’s latest models struggled; notably, o3-mini solved only 2% of the tasks.

For a technical deep-dive on how Kotlin-bench was curated and how we scaled thousands of Android containers in parallel for evaluation, read more here:

- https://firebender.com/blog/kotlin-bench

I also open sourced all the code and datasets:

- https://github.com/firebenders/Kotlin-bench

- https://huggingface.co/datasets/firebenders/Kotlin-Bench

Thanks for reading! I'd love feedback, questions, or suggestions on how to improve Kotlin-bench further!

Comments URL: https://news.ycombinator.com/item?id=43587722

Points: 1

# Comments: 0