Natural Language Processing news and analysis on Web Pulse: 5621 curated articles tracking the latest Natural Language Processing developments, tools, and research, updated continuously from vetted sources.
A writer who used LLMs to polish a math blog later recognized distinct "AI smells" — recurring patterns like excessive punchlines, consecutive short sentences, and specific formatting choices — appearing across the inter…
The University of Central Oklahoma's Forensic Science Institute is using Google's NotebookLM artificial intelligence tool to reduce the time required to analyze complex criminal case documents and construct timelines fro…
A developer built a human-in-the-loop AI workflow automation system using Make, FastAPI, OpenAI, and Monday CRM to reduce repetitive manual review work while keeping the process controlled and traceable. The system proce…
A developer measured the token cost of feeding raw HTML versus extracted text to large language models across 10 real web pages, finding a median token multiplier of 7.4×. The spread ranged from 1.1× on a stripped-down t…
Apple will unveil a redesigned Siri with ChatGPT-style conversational AI features at its Worldwide Developers Conference on June 8, powered by Google's Gemini technology. The overhaul, arriving with iOS 27 and macOS 27, …
Apple is reportedly redesigning Siri with a ChatGPT-style conversational interface, including a dedicated Siri app and Dynamic Island integration, according to leaked renders and reporting. The overhaul would replace the…
Sesame, the conversational AI startup co-founded by Oculus founders, launched its iOS app on Thursday, offering a public preview of AI agents designed to speak more naturally by running parallel searches and weaving resu…
Ahead of Apple's Worldwide Developers Conference in June, Bloomberg published leaked renders of Apple's planned AI upgrade for iPhone, revealing a new Siri app designed to rival ChatGPT and other AI chatbots. The images …
A developer built Provenant, a system that generates human-readable wiki pages for every file and module in a codebase, then searches those pages instead of raw source code. On SWE-bench Verified across 500 GitHub issues…
Hippo expanded its voice-based AI customer service representative, Hannah, from an after-hours tool into the first point of contact for all inbound customer calls. According to Insurance Innovation Reporter, Hannah has h…
New renders from Bloomberg show Apple's iOS 27 Siri overhaul featuring a pill-shaped chat bubble that emerges from the Dynamic Island and a standalone Siri app resembling ChatGPT's interface. The redesign, which includes…
A developer built Issue AI Agent, a 500-line TypeScript GitHub Action that automatically triages repository issues in about eight seconds. The bot classifies each issue, detects duplicates, applies labels, and posts cont…
A developer has built a privacy-first Tampermonkey userscript called ChatGPT Long Conversation Helper that adds collapse and expand controls for individual messages in long ChatGPT conversations. The tool, available on G…
Bloomberg's Mark Gurman shared re-created screenshots of iOS 27's new Siri app and "Search or Ask" feature, revealing a dark-themed interface with text and voice modes, conversation history, and integration with Apple In…
A developer has released `whenis`, an open-source tool that moves date interpretation out of LLMs and into deterministic parsing, addressing the common problem of AI agents hallucinating weekday-to-date mappings and rang…
Ekorbia v0.2 introduces a comparison-chat mode that runs two to three local large language models against the same prompt in parallel. Testing Gemma 4, IBM Granite 4.1, and Qwen 3.5 on a 32 GB M1 Max MacBook Pro revealed…
RankOps has launched a City GEO Hub for the Charlotte, NC metro market, designed to optimize for AI search visibility by combining entity clarity, structured data, and local context. The page targets the city's fast-grow…
An iPhone user who tested Google Gemini through Android Auto in a Toyota Camry found the AI assistant outperformed Siri for in-car tasks like sending emails, getting restaurant recommendations, and answering complex ques…
Five frontier large language models disagreed on 67% of 1,000 real-world fact-check claims submitted by users to a verification platform, according to a new analysis. The panel of top AI models failed to reach a unanimou…
A developer built an AI tool called RoastMyEmail that scored over 18,000 cold email drafts and found the average first draft scores just 32 out of 100. The analysis revealed that most cold emails fail for the same system…