3
points
kure256
Joined 17 karma
- The goal is to adapt your HTML to be easily readable by crawlers, making it more likely to be cited by AI assistants such as ChatGPT, Claude, Gemini, or Perplexity, instead of relying solely on traditional search engines.
https://github.com/ai-first-guides/first.ai?tab=readme-ov-fi...
- - https://github.com/ai-first-guides/first.ai?tab=readme-ov-fi...
AI-first web design is based on a simple observation: more and more queries go directly to AI assistants instead of traditional search engines. If you want AI systems to understand, trust, and cite your content, the web must be written for them — not only for the visual browser. Just like SEO once made websites visible to Google, AI-first structure increases the chance that your website becomes the source AI systems quote.
- A small additional note for context:
I’m not arguing that “LLMs will replace browsing” in some absolute way — but it is observable that for many users, the entry point for information is shifting from search → assistant. When you actually inspect how models consume real websites today, the results are pretty uneven:
pages with clean HTML and predictable structure get parsed reliably
JSON-LD is used surprisingly often (but only if it’s correct and minimal)
heavy client-side rendering breaks extraction more than people expect
semantic markup still beats any “AI-enabled” tool by a mile
models hallucinate less when the source has clear hierarchy and meaning
This project isn’t trying to reinvent SEO — it’s more like exploring the minimum structural guarantees that make an LLM treat a page as a trustworthy, cite-able source instead of ignoring it or misreading it.
If anyone here has done experiments with:
how GPT, Claude, Gemini, Llama, etc. read arbitrary web pages
failure cases in parsing / hallucination caused by layout
the effect of metadata vs full-text signal
or even prompt strategies for web ingestion
…I’d genuinely love to compare notes.
- 1 point
- More and more people now start with an AI assistant instead of traditional browsing — not because they love AI everywhere, but because it’s simply faster than navigating websites. The shift is already visible: assistants can surface structured information directly, and they’re beginning to prioritize citations, so sources that are clear and machine-readable get more visibility.
If the web doesn’t adapt, a lot of high-quality content will slowly disappear from the “AI layer” of discovery.
We’re trying to document this shift here: https://github.com/ai-first-guides/first.ai/blob/main/docs/i...
- 1 point
- More users are now getting information through AI assistants (ChatGPT, Gemini, Copilot) instead of traditional browsing. This project proposes a set of open, practical guidelines for building websites that AI systems can actually understand, index, and cite — similar to how SEO defined rules for search engines.
It’s not a framework — just a minimal, evolving set of best practices aimed at making the web machine-readable again.
Repo (docs): https://github.com/ai-first-guides/first.ai/blob/main/docs/i...
Feedback, Contribution and Stars from HN is welcome.
- 3 points
- We’ve been experimenting with multi-cluster failover for Kubernetes workloads, and one open-source project that actually works really well is k8gb .
It acts as a GSLB controller inside Kubernetes — doing DNS-level health checks, region awareness, and automatic failover between clusters when one goes down.
It integrates with ExternalDNS and supports multiple DNS providers (Infoblox, Route53, Azure DNS, NS1, etc.), so it can handle failover across both on-prem and cloud clusters.
It’s not a silver bullet for every architecture, but it’s one of the few OSS projects that make multi-region failover actually manageable in practice.
- A Global Service Load Balancing solution with a focus on having cloud native qualities and work natively in a Kubernetes context.
``` apiVersion: k8gb.absa.oss/v1beta1 kind: Gslb metadata: name: test-gslb-failover namespace: test-gslb spec: resourceRef: apiVersion: networking.k8s.io/v1 kind: Ingress matchLabels: # ingresses.networking.k8s.io resource selector app: test-gslb-failover strategy: type: failover # Global load balancing strategy primaryGeoTag: eu-west-1 # Primary cluster geo tag ```
- 2 points
- 1 point
- 1 point
- Thats Excellent! I did not know that you can make dye from mushrooms - I come from central Europe and we do go mushrooming - we call that mushroom hunting. My family goes mushrooming in the woods with the kids and spends time together. It's a good way to spend time. Afterwards, the mushrooms are used to prepare food.
Apart from food, I know that the poor people used to make hats and other things out of mushrooms. I read somewhere that they also used them as building material.
- 1 point
- Location: Prague, Czechia (EU) Remote: Yes Willing to relocate: No Technologies: paintshop, photoshop, illustrator, indesign , cinema4d, sketchup, after effects.
Position: junior graphics Résumé/CV: https://www.alenakuritka.com/ Email: advert.kuritka@gmail.com
I like to learn new things, I’m happy to accept new offers in EU.
- 28 points
The goal is to make content readable by anything. As more users access information through systems like ChatGPT instead of visiting websites directly, content that isn’t easily interpreted by AI crawlers risks becoming effectively invisible.
see: https://github.com/ai-first-guides