Reading up on search-engine
63 deep · digging since nov 21, 25
- Cloudflare gives OpenAI network signals covering 20% of the web
Cloudflare and OpenAI launched a research pilot that feeds Cloudflare's real‑time network freshness and traffic data into OpenAI's search system to speed up and improve AI web indexing.
- Boringlaunch - Increase Your SEO Score In 30 Days
Boringlaunch offers a service that submits AI startups to over 100 free platforms within seven days, promising improved domain rating and SEO scores within thirty days.
- Bing Webmaster Tools
Bing Webmaster Tools provides free SEO/GEO tools, backlinks analysis, keyword research, and site scanning to help webmasters improve site performance in Bing search results.
- Show HN: Bible as RAG Database
A user-built semantic search tool for the Bible using RAG, with commenters discussing its accuracy, features, and the theological implications of AI-interpreted scripture.
- Show HN: I made Google Trends for Hacker News by indexing 18 years of comments
Hacker Trends indexes 18 years of Hacker News comments to let users compare term frequency over time, built on Upstash Redis Search.
- A website that lists websites to submit your website to
Hacker News commenters discuss a website listing 50 directories for submitting websites, debating the declining SEO value and risk of spam.
- Google tests sending Chrome users straight into AI Mode
Google tests a Chrome flag that redirects omnibox searches to AI Mode but insists it is just an exploration with no plans to change the default behavior.
- Companies Are Using Reddit to Manipulate ChatGPT and Google AI Search
Peptide companies are spamming Reddit's biohackers subreddit to manipulate AI chatbots and search engines like ChatGPT and Google into recommending their products.
- Google offers opt-out of “AI” search results for websites, promises it won’t affect regular search rankings – OSnews
Google adds a Search Console toggle allowing websites to opt out of AI Overviews and other generative AI search features, promising no impact on regular rankings.
- Rethinking Search as Code Generation
Perplexity introduces Search as Code (SaC), an architecture that exposes search stack components as programmable SDK primitives for AI agents to compose via generated Python code.
- DuckDuckGo search saw 28% more visits after Google said people love AI mode
DuckDuckGo saw a 28% increase in visits after Google promoted its AI mode, sparking polarized reactions on Hacker News.
- Sundar Pichai Understands Why People Are Anxious About A.I.
Sundar Pichai acknowledges public anxiety about AI while discussing Google's integration of AI agents into Search and his advice for graduates.
- Google Pushes AI-Generated Ads Further Into Search Results - WSJ
Google is testing new ad formats in standard search and its AI Mode to convert AI features into ad revenue as Meta threatens to overtake its digital ad dominance.
- How Google plans to win the AI war
Google is pursuing a dual strategy of aggressively integrating AI into its products while protecting its core advertising and cloud revenue streams.
- Apparently Google hates us now
HN commenters argue that Google's services have degraded due to enshittification, driving users to alternatives like Kagi, Yandex, and DuckDuckGo.
- Google changes its search box
Google's shift to AI-generated search summaries threatens the ad-revenue web model by extracting content without sending traffic back to publishers.
- Google Changes Its Search Box for the First Time in 25 Years
Google overhauls its search box with Gemini AI to handle longer queries, adds a video-generation tool, and simplifies online shopping.
- Agentic search models
Smaller, domain-tuned LLMs trained specifically for search orchestration can replace complex monolithic retrieval pipelines by intelligently controlling simpler retrieval primitives.
- Ask.com has closed
Ask.com, the iconic question-based search engine, shut down on May 1, 2026, after 30 years of operation.
- Five Ways A.I. Search Beats an Old-School Google Search
Google's AI search outperforms traditional search for practical tasks like grocery selection and scam detection, but remains unreliable for celebrity news.
- How We Improved Agentic Search
Ranking search results to surface definitions before tests and vendor code improves first-query retrieval and helps coding agents find the right files sooner, while raw speed gains alone barely reduce end-to-end runtime.
- Google Search AI Mode Gets 'Expert Advice' From Reddit and Social Media - MacRumors
Google is adding a 'preview of perspectives' to AI search results, sourcing content from Reddit and social media and sometimes labeling it as 'Expert Advice'.
- RSS Feeds Send Me More Traffic Than Google – Terence Eden’s Blog
Blogger Terence Eden finds RSS feeds and email subscribers drive more traffic to his personal site than Google Search over a 28-day period.
- Ex-Twitter CEO’s AI Startup Raises Funds at $2 Billion Valuation - WSJ
Parallel Web Systems, founded by ex-Twitter CEO Parag Agrawal, raised $100M in Series B funding at a $2 billion valuation to build web search for AI agents.
- Can agents replace the search stack?
Using a basic BM25 or e5 retriever with an LLM agent can achieve 0.289→0.453 NDCG on Amazon ESCI by reasoning over queries, but this approach fails on passage retrieval where the embedding model already knows best.
- Google upgrades AI Mode in the Chrome browser
Google updates AI Mode in Chrome to open web pages side-by-side with search, enabling contextual follow-up questions and cross-tab search across tabs, images, and files.
- The “small web” is bigger than you might think
A Hacker News discussion argues the small web is defined by a non-commercial mindset, not size, and remains sizable yet buried by algorithm-driven search engines.
- Fast regex search: indexing text for agent tools
Cursor is building local indexes for regular expression search to reduce agent wait times from 15-second ripgrep runs in large monorepos.
- Why your next mobile app is probably headless - Tuan-Anh Tran
As AI assistants answer queries and complete tasks without opening vendor apps, mobile apps become headless backends, and metrics like traffic and sessions decline long before installs.
- I'm losing the SEO battle for my own open source project
An open-source project maintainer loses search-engine ranking to a copycat .net site, and HN commenters debate SEO fixes, trademark strategies, and the broader decay of Google search quality.
- Google expert explains AI Mode in Search’s query fan-out method
Google's AI Mode uses a query fan-out technique to simultaneously search multiple objects in one image, delivering cohesive results via Gemini models.
- Nano Banana 2: Google’s latest AI image generation model
Google DeepMind launches Nano Banana 2, combining advanced image generation features with Gemini Flash's speed for rapid editing and iteration.
- Reddit says 80 million people now use its search weekly
Reddit reported 80 million weekly search users after merging its core search with AI-powered Reddit Answers, positioning itself as an end-to-end discovery engine.
- Everything – Locate files and folders by name instantly
Everything is a Windows utility that instantly locates files and folders by name by indexing the NTFS file table into memory for near-instantaneous search results.
- Securely indexing large codebases
Cursor cuts time-to-first-query from hours to seconds on large codebases by securely reusing a teammate's prebuilt index via Merkle tree-based syncing and access proofs.
- AI Mode in Google Search and AI Overviews get Gemini upgrades
Google Search globally rolls out Gemini 3 for AI Overviews, enabling follow-up questions to start conversational AI Mode.
- Building docfind: Fast Client-Side Search with Rust and WebAssembly
Docfind uses Rust, WebAssembly, and FSTs to create a compact client-side search engine that runs entirely in the browser, with Copilot accelerating development.
- Exa AI Research Blog | Semantic Search & Neural Network Search Engine
Exa built exa-d, a data framework using typed columns, declarative dependencies, and surgical fragment updates to manage a real-time web index at scale.
- MSN
MSN is a Microsoft-owned web portal that provides news, email, and search services.
- Attention Required!
seats.aero is a fast search engine for finding award flights using points, offering free searches and alerts with a PRO tier for extended features.
- Let Me Droid That For You: What 780K Agent Searches Reveal
Analysis of 780K AI agent web searches reveals a dominant pattern: exploratory search followed by specific fetch, with 75% of queries related to software development.
- How Google Got Its Groove Back and Edged Ahead of OpenAI - WSJ
Google regained the AI lead over OpenAI by combining deep research, custom hardware, and leadership changes, culminating in Gemini's rise and a major search overhaul.
- Google broke my heart
An author shares his frustrating experience trying to remove pirated copies of his books from Google search results, only to face a broken DMCA verification process.
- AI Mode, Content & Search Index – DEJAN
Tests show Google's AI Mode retrieves page content from a proprietary store rather than the live web search index, contradicting expectations.
- New Google web ecosystem tools and partnerships
Google launches Preferred Sources globally, highlights news subscriptions in the Gemini app, and pilots AI partnership programs with news publishers to drive engaged audiences.
- GitHub - tobi/qmd: mini cli search engine for your docs, knowledge bases, meeting notes, whatever. Tracking current sota approaches while being all local
QMD is a local hybrid search engine for markdown documents that combines BM25, vector search, and LLM reranking, with CLI, MCP, and SDK interfaces.
- Google adding seamless transition from AI Overviews to AI Mode
Google is testing a seamless transition from AI Overviews to AI Mode on mobile, and expanding Gemini 3 Pro to 120 countries.
- 28M Hacker News comments as vector embedding search dataset
ClickHouse provides a 28.74-million Hacker News posting dataset with vector embeddings, enabling large-scale vector search application development and testing.
- Search | The Moat of the Search Index - by FD - Robonomics
The author argues that the traditional search index moat is largely gone because AI assistants like ChatGPT use multi-hop reasoning across multiple sources, hiding individual search failures.
- Just a moment...
Perplexity introduces memory features for its AI assistant, allowing it to remember user preferences and conversations across models for more personalized responses.
- Thread by @SashaKaletsky on Thread Reader App – Thread Reader App
Perplexity's global app downloads dropped 80% over six weeks, suggesting its earlier growth was largely fueled by paid marketing rather than organic user acquisition.
- Kagi Bloopers – Search Results Gone Wrong
Kagi compiles a collection of humorous search results that went hilariously
Takes
It works! AI labs started indexing, training, and serving my new *.md files ✨ Right after pushing to prod, I asked ChatGPT to check {{startup_markdown_URL}}, and it couldn't because of "cache missed" or "400 timeout fetching". So I added all 8,831 URLs to the sitemap.xml and added the following headers: "Content-Type": "text/plain; charset=utf-8" "Content-Disposition": "inline" I can't tell if this helped, but 24 hours later, ChatGPT was able to fetch them properly 😊 Maybe AI assistants need the page to be indexed first, before being able to crawl it?
@marclou
introducing, the @stripe directory: from the cli, people and agents can search for, and pay, businesses on stripe. > stripe search "serverless postgres database" - payments between two stripe users are free - includes mpp and projects services - join by creating a profile =>
@jeff_weinstein
anyone built a reliable system/solution for indexing and searching threads across claude code, codex, cursor..etc? All the built-in ones are not that great.
@shadcn
If you need AI to do a search for you in the real world, ds4-agent is basically SOTA, because it can access the web sites without any limitations given that it uses your local Chrome browser (no, not in headless mode, that's the trick...), and DeepSeek v4 is great at search.
@antirez
while i've had SEO success on some sites, others...not so much. 🫣
@Shpigford
episode 04 of my show life of an indie maker - SEO is dead 💀
@tibo_maker
SEO folks: What's the current thought on submitting to directories for new products/domains to increase DR (especially when DR is <10).
@Shpigford
Exa is what I trust for all my agents. We use it at YC. We use it in all my OpenClaw and Hermes Agents. There is no other option that is as fast, as reliable, and as complete. When your agents need to search the web, accept no substitutes.
@garrytan
Introducing Exa for Claude: Give Claude access to billions of websites, docs, papers, people, companies, and more. Exa turns Claude into a recruiter, a salesperson, a researcher, a reliable engineer, all with one plugin. Run in terminal: claude plugin marketplace update claude-plugins-official && claude plugin i exa@claude-plugins-official && claude "call mcp__plugin_exa_exa__authenticate (a deferred tool available to you, guaranteed)"
@ExaAILabs
We’ve post trained a model on top of Qwen that achieves Pareto optimality on accuracy-cost curves. Unlike our previous post trained models, this model has been trained to be good at search and tool calls simultaneously, allowing us to unify the tool call router and summarization together in one model. The resulting model performs better than GPT and Sonnet in terms of cost efficiency to serve daily Perplexity queries in production. The production model runs on our own inference platform. We’re already serving a significant chunk of our daily traffic with this model and intend to have it serve all of default traffic pretty soon. More research to follow soon on models we’re training and deploying for Comet and Computer.
@AravSrinivas
Today we are launching Orion Browser v1.0https://t.co/5gMP24nT3LIt has been quite a journey for the last 6 years building not just a search engine so good that people would pay for it, but a web browser too.Thank you all loyal users, supporters and customers. And thank… pic.twitter.com/dHcU1qwYXn
@vladquant