Recent submissions

Comment · 3 months ago
How do you run this phases and parallelization on each?

Via just ... "prompting it"?

Or do you use any tool in the middle to ensure this agent architecture?

Just curious if there is any workflow-like tool in the middle that is helping.

Comment · 3 months ago
Not yet in Antigravity CLI 1.1.5 for Antigravity Business
Comment · 4 months ago
Fugu Ultra <https://console.sakana.ai/models#fugu-ultra> sounds similar to GPT-5.5 Pro or Gemini 3.1 Deep Think .

Is there any official source that could confirms if Fable (or Mythos) is parallelized test-time compute (like GPT 5.5 Pro) or sparse Mixture-of-Experts (MoE) transformer combined with a multi-agent, inference-time compute scaling architecture (Gemini 3.1 Deep Think)?

Comment · 4 months ago
Comment · 6 months ago
are you familiar with worktree? or are you asking for something more specific or elaborated?
Comment · 7 months ago
Comment · 7 months ago
made me remember this article

<https://adamdrake.com/command-line-tools-can-be-235x-faster-...>

  Command-line Tools can be 235x Faster than your Hadoop Cluster (2014)

  Conclusion: Hopefully this has illustrated some points about using and abusing tools like Hadoop for data processing tasks that can better be accomplished on a single machine with simple shell commands and tools.
Comment · 8 months ago
Would you mind sharing or pointing me to any links that explain how you've set up Zig to work effectively with LLMs using agentic features?

I was thinking that downloading the full official documentation, separated by sections inside the repository someone is working on with Zig, could be useful, but maybe there are more optimal ways to approach this.

Thanks.

Comment · 9 months ago
..: Instant QR-based Account-to-Account (A2A) Payment System
Comment · 9 months ago
Six years old might be too early to incentivize screen use, though I'd encourage you to research this yourself rather than taking internet comments at face value.

That said, I think there's a distinction between screens and computing itself. You could introduce her to computing power through voice interfaces: a smart speaker connected to an LLM could let her search, learn, and interact with information hands-free. You'd have control over the system prompts for safety, and could whitelist reliable sources for her queries.

Yes, visual information density is higher than audio, but the downsides of early screen exposure might outweigh that efficiency gain. Voice-first computing could be a middle ground, she gets to explore what computers can do without the attention/addiction patterns that screens introduce.

Just one perspective, obviously. Worth doing your own research on the developmental tradeoffs.

Comment · 9 months ago
When I have that need, this is what I do.

I use the info from here as context:

https://github.com/InteractionDesignFoundation/add-event-to-...

I load it via a custom system prompt or skill, then pass my specific need, thought, or anything I want to remember in the future, including events with a certain frequency, as part of the prompt.

From that, I render the URLs needed to create the events directly in my personal calendar via the browser. This part could probably be automated better, but honestly I'm lazy sometimes.

And that's it.

I also try not to overdo it with high-frequency reminders, since that tends to de-incentivize actually using your personal calendar, which kind of defeats the purpose.

On top of that, Telegram 'Saved Messages' with the reminder feature is really useful. The native app makes it very fast to navigate. Obviously not as fast as searching local plain text files, but definitely faster than WhatsApp.

Comment · 9 months ago
Is that actually required?

I'm not at my computer right now, but as far as I remember most LinkedIn posts are viewable without logging in if you use the direct URLs, for example:

https://www.linkedin.com/feed/update/urn:li:ugcPost:<post_id>

https://www.linkedin.com/feed/update/urn:li:activity:<activity_id>

Unless something changed recently, those links should work anonymously.

Comment · 10 months ago
Comment · 10 months ago
I have a similar workflow and I'd like to know how sophisticate is your requirement for annotations, because I solve that part kind of smoothly with Excalidraw
Comment · 10 months ago

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton