Natural Language Processing news and analysis on Web Pulse: 5440 curated articles tracking the latest Natural Language Processing developments, tools, and research, updated continuously from vetted sources.
Proto Hologram, a Los Angeles company, has installed nine holographic display boxes at Boston Logan International Airport featuring a lifesize AI character named Amelia that answers traveler questions in several language…
XAI launched Grok Voice on fal.ai, giving developers a real-time speech-to-speech API with response latency around 0.70 seconds, support for over 25 languages, 26 voice options, and audio cloning. Pricing is set at $0.00…
ByteLex reported that a tokenizer-free coordinate map built from 11 tokenizer vocabularies over 3-byte windows predicted its 237M-parameter byte-level language model's per-word errors with a Spearman correlation of −0.98…
Rijul, a developer building the AI code review tool LiveReview, outlined the difference between autoregressive and diffusion-based text generation. Autoregressive models, used by most deployed language models today, gene…
JanAI Technology Private Limited launched its first platform version across nine JanAI Cafes, with seven cafes in Mysuru district and Prayagraj district, aiming to bring AI-driven services to 100 million rural Indian hou…
A developer built MedScribe, an offline clinical documentation pipeline that pairs FasterWhisper ASR with LLM-based medical term extraction, after finding that standard confidence scores cannot reliably gate auto-correct…
Shenava-Koochik, a 114M-parameter Persian streaming ASR model from Reza2kn's Shenava collection, is not the compute bottleneck in Persian speech pipelines; the LLM decision/intent stage is where failures occur, according…
A developer outlined a grounded Retrieval-Augmented Generation (RAG) pipeline that pairs local vector stores with strict context prompts to curb LLM hallucinations. The approach chunks documents with sliding-window token…
Retrieval-augmented generation (RAG) lets LLM applications answer questions about private company documents by retrieving relevant passages rather than relying on the model's built-in knowledge, according to a technical …
Slack published a 2026 buyer's guide to enterprise AI chatbot solutions, listing Slackbot and Slack AI, Microsoft Copilot for Microsoft 365, Gemini Enterprise, ChatGPT Enterprise, Claude Enterprise, Notion Enterprise, Za…
A Pew Research Center study of 68,879 Google searches by 900 U.S. adults in March 2025 found that users clicked a traditional search result in 8% of visits where an AI Overview appeared, compared with 15% of visits witho…
Google released new Gemini Live models in the Gemini API and Google AI Studio, including Gemini 3.8 Live and 3.8 Live Extended Thinking, which the company says rank first on Artificial Analysis' Speech-to-Speech leaderbo…
Google released Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking on September 15, two voice models that switch languages mid-sentence across 97 languages, process near real-time video, and make API calls in the back…
A developer published a tutorial demonstrating sentence-window retrieval-augmented generation (RAG), a technique that indexes individual sentences for precise matching and then expands each hit into a surrounding window …
KDnuggets published a cheat sheet on Scikit-LLM, a library that wraps language models in the scikit-learn estimator API so they drop natively into a Pipeline or cross-validation loop. The cheat sheet covers estimators in…
Speechify released Simba 3.2 and Simba 3.0, text-to-speech models that cut time-to-first-audio to roughly 100ms on Coval's benchmark while keeping the same voices, quality, and per-character pricing. The company said the…
Google Research introduced R4T, an RL-trained diffusion retriever that optimizes fan-out query sets for diversity and coverage, achieving 12x to 20x latency speedups over autoregressive language models. The 53.9M-paramet…
Pinecone, a fully managed vector database launched as a public beta in January 2021, is most commonly used today for retrieval-augmented generation (RAG) and agent memory, according to the company's positioning as an AI …
A scan of 10,126 pages of real marketing copy across 103 sites, run between 2026-08-14 and 2026-09-14, found a median site score of 90/100 but at least one page scoring under 60 on 17% of sites, according to SiteTell. Th…
A developer measured token usage across 10 web pages and found that raw HTML consumed 3 to 24 times more tokens than the same pages converted to Markdown, with one product page arriving as 267,361 tokens of which only 3,…