Feed aggregator

Article URL: https://josephg.com/blog/crdts-are-the-future/

Comments URL: https://news.ycombinator.com/item?id=44156626

Points: 1

# Comments: 0

Article URL: https://secret.club/2025/06/02/hypervisors-for-memory-introspection-and-reverse-engineering.html

Comments URL: https://news.ycombinator.com/item?id=44156625

Points: 1

# Comments: 0

Article URL: https://makeincoimbatore.substack.com/p/the-case-for-coimbatore

Comments URL: https://news.ycombinator.com/item?id=44156622

Points: 1

# Comments: 1

Noteworthy stories that might have slipped under the radar: simple PoC code released for Fortinet zero-day, OpenAI O3 disobeys shutdown orders, source code of SilverRAT emerges online.

The post In Other News: PoC for Fortinet Bug, AI Model Subverts Shutdown, RAT Source Code Leaked appeared first on SecurityWeek.

Article URL: https://www.theguardian.com/sport/2025/jun/01/magnus-carlsen-punches-table-norway-chess-gukesh

Comments URL: https://news.ycombinator.com/item?id=44156602

Points: 1

# Comments: 0

Article URL: https://scitechdaily.com/mits-sodium-fuel-cell-powers-planes-captures-carbon-and-outruns-batteries/

Comments URL: https://news.ycombinator.com/item?id=44156592

Points: 1

# Comments: 0

Article URL: https://revontulet.dev/p/2025-ai-first/

Comments URL: https://news.ycombinator.com/item?id=44156570

Points: 2

# Comments: 0

Struggling to find the perfect café in Rishikesh that serves great food, amazing coffee, and a peaceful vibe? EazyDiner has got you covered! Explore the best cafes in Rishikesh, offering stunning Ganges views, organic meals, and cozy spots to relax after a long day of exploring or yoga. Whether you want a quiet escape or a lively hangout, find the right place with exclusive deals. Book now on EazyDiner and enjoy the best café experiences hassle-free!

Explore on: https://www.eazydiner.com/rishikesh/restaurants/cafe

Comments URL: https://news.ycombinator.com/item?id=44156561

Points: 1

# Comments: 0

Article URL: https://tuist.dev/blog/2025/05/13/vibe-xcoding

Comments URL: https://news.ycombinator.com/item?id=44156555

Points: 1

# Comments: 0

Article URL: https://www.youtube.com/watch?v=y8eMABF1hD8

Comments URL: https://news.ycombinator.com/item?id=44156549

Points: 1

# Comments: 0

Article URL: https://veryl-lang.org/blog/annoucing-veryl-0-16-1/

Comments URL: https://news.ycombinator.com/item?id=44156543

Points: 1

# Comments: 0

Article URL: https://www.theregister.com/2025/05/30/meta_is_now_a_defense/

Comments URL: https://news.ycombinator.com/item?id=44156531

Points: 2

# Comments: 1

Article URL: https://www.theneuron.ai/explainer-articles/the-ai-jobs-reality-check-whats-actually-happening

Comments URL: https://news.ycombinator.com/item?id=44156526

Points: 1

# Comments: 0

Article URL: https://staysaasy.com/managementya/2024/01/07/performance-reviews.html

Comments URL: https://news.ycombinator.com/item?id=44156522

Points: 2

# Comments: 0

Article URL: https://mortalityreminder.com/

Comments URL: https://news.ycombinator.com/item?id=44156513

Points: 1

# Comments: 0

Article URL: https://github.com/beehive-lab/GPULlama3.java

Comments URL: https://news.ycombinator.com/item?id=44156512

Points: 1

# Comments: 0

Article URL: https://oxfordamerican.org/oa-now/inside-rice-university-s-2025-ufo-conference

Comments URL: https://news.ycombinator.com/item?id=44156500

Points: 2

# Comments: 1

Article URL: https://james.cridland.net/blog/2025/local-radio-future/

Comments URL: https://news.ycombinator.com/item?id=44156483

Points: 2

# Comments: 0

I built a system that lets LLMs automatically learn and improve problem-solving strategies over time, inspired by Andrej Karpathy's idea of a "third paradigm" for LLM learning.

The basic idea: instead of using static system prompts, the LLM builds up a database of strategies that actually work for different problem types. When you give it a new problem, it selects the most relevant strategies, applies them, then evaluates how well they worked and refines them.

For example, after seeing enough word problems, it learned this strategy:

1) Read carefully and identify unknowns,

2) Define variables with units,

3) Write equations,

4) Solve step-by-step,

5) Verify the answer.

All strategies are stored as human-readable JSON that you can inspect and edit.

I tested it on math benchmarks and saw decent improvements - 8.6% better on Arena Hard, 6.67% on AIME24. After 500 queries, the system had created 129 strategies and refined 97 of them.

The implementation is an open-source plugin for optillm (our inference optimization proxy). It works with any OpenAI-compatible API - you just add "spl-" to your model name. Has two modes: inference-only (uses existing strategies) and learning mode (creates and refines strategies).

What's interesting is that it bridges the gap between the sophisticated system prompts that production AI uses and the basic prompts most of us work with. Your model literally gets better at the types of problems you throw at it.

Built it because I noticed ChatGPT, Claude etc. have incredibly detailed system prompts with problem-solving frameworks, but most developers use basic prompts and miss out on those performance gains. The approach is inspired by Andrej Karpathy's tweet about a "third paradigm" for LLM learning beyond just pretraining and fine-tuning: https://x.com/karpathy/status/1921368644069765486

The strategies are completely transparent - you can see exactly what the system learned and why it's making certain decisions. No black box learning.

https://github.com/codelion/optillm/tree/main/optillm/plugin...

Would love feedback on the approach. Has anyone else experimented with LLMs learning from their own experience?

Comments URL: https://news.ycombinator.com/item?id=44156467

Points: 3

# Comments: 1

Hi HN,

I’ve been building Onboardzy, a tool that lets developers and product teams create, test, and update mobile onboarding flows without having to ship a new app version or wait for app store reviews.

Since sharing the MVP here a while back, I’ve added: • Split testing – run A/B tests to find the best-performing onboarding flows • Integrated analytics – see average completion rates and slide-by-slide drop-offs

Onboardzy is live on Product Hunt today, and I’d love to hear your thoughts or any feedback on how to make it even better for developers.

Here’s the link if you’d like to check it out: https://www.producthunt.com/products/onboardzy

Thanks for taking a look! I’ll be around in the comments if you have questions or suggestions.

Armin

Comments URL: https://news.ycombinator.com/item?id=44156387

Points: 1

# Comments: 0