AI Search news and analysis on Web Pulse: 728 curated articles tracking the latest AI Search developments, tools, and research, updated continuously from vetted sources.
Manticore Search switched columnar vector access to memory mapping, raising KNN rescoring throughput by 2.57x–3.13x and reaching 85–92% of row-wise performance, the company reported. In a DBpedia benchmark on a 16-core A…
Reddit will shut off free public API and RSS access in four stages — no new public API requests on October 31, 2026, RSS feeds dark on November 13, unregistered third-party apps losing access on January 12, 2027, and the…
A developer detailed how edge security rules and web application firewalls—not robots.txt—are the usual cause of AI crawlers like GPTBot being silently blocked from websites. The writeup explains that Cloudflare's Super …
A developer released SCM 0.2.4, a local-first AI search app for macOS that indexes every photo and every frame of video in watched folders using on-device vision models, OCR via Tesseract, and dialogue search via Whisper…
A difference-in-differences study of Google's AI Overviews (AIO) found that default AIO availability reduced English Wikipedia's external search traffic by 5.45% versus German and 4.82% versus French, according to the ar…
A developer published a 30-minute manual audit method for checking whether AI assistants such as ChatGPT, Perplexity and Gemini recommend a given brand. The process has users write 15 to 20 real customer questions, run t…
Dejan.ai's app.dejan.ai tracks canonical core entities rather than individual prompts to score brand visibility across Google, OpenAI and Anthropic models, running one fixed probe prompt per entity with web search on and…
The Wanloria team built a free bilingual (Spanish/English) travel web app covering verified places in 71 cities, 1-3 day itineraries, free-attraction lists and signposted hiking trails, largely by pair-programming with C…
A developer at the agency Pagepro built Cross-Check, an agent that reads the company's website through a Sanity Knowledge Base and compares it against 87 real buyer questions and 65 sales answers, filing gaps, ambiguitie…
A developer built InsightTrack, a tool that combines analytics traffic data with SerpApi SERP data to explain why a page's traffic changed, using deterministic attribution rules rather than letting an AI guess. The proje…
A developer published a guide to Valyu's DeepResearch API, detailing five ways to run multi-step research tasks that produce deliverables such as Markdown, PDF, schema-based JSON or TOON reports, plus spreadsheets, docum…
Astra Intelligence released Web2MD, an open-source MCP (Model Context Protocol) server that converts any public URL into clean Markdown stripped of ads, navigation, and boilerplate for LLM consumption. The server exposes…
A University of Copenhagen study of 27 large language models found that GPT-5 produced answers 18.5% less diverse than Google Search results, while the average chatbot showed 19% less diversity, based on over 70 million …
Turbopuffer, the vector search layer behind Cursor and Notion, announced in a September 30 blog post that it is removing the vector index as the primary data structure in its next major version, citing storage amplificat…
A developer released GhostBench, a benchmark that tested whether AI models recommend shut-down SaaS vendors, finding that 19 of 20 dead products' domains still return HTTP 200 and that Claude Haiku 4.5 presented 8 of 20 …
Top AI models answer roughly 1 in 4 specific-fact questions incorrectly when they cannot search, according to Artificial Analysis data cited by Randy Olson. The AA-Omniscience dataset covers 6,000 questions on specific f…
A Google Cloud employee ran a controlled experiment with 310 prompts to test whether shaping search results could steer AI coding agents toward Google Cloud products, after an Armature study of 16,893 coding-agent sessio…
Arda launched a marketing agent product that tracks and improves brand visibility across Google and AI answer engines including ChatGPT, Claude, Perplexity, and Gemini, publishing CMS-ready pages, on-page fixes, comparis…
Kapa co-founder and CTO Finn Bauer said the company's Kapa Deep retrieval system scored 0.65 on Company Knowledge Bench, a private benchmark published October 2nd built from 1,000 real production queries, edging out Kapa…
A test of 75 answers from ChatGPT, Claude, Gemini, Perplexity and Grok found that Next.js, Supabase, Vercel and Stripe were named together as the default app stack in 51 of 75 responses (68%), and in 42 of 45 answers whe…