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.
Amazon Web Services announced the general availability of TwelveLabs Marengo Embed 3.0 as an embedding model in Amazon Bedrock Knowledge Bases, enabling natural language search across video, audio, and image content. Mar…
OpenAI launched GPT-Live-1 in its API, a full-duplex voice model that listens and speaks simultaneously and costs $0.05 per minute, with backend model and agent harness fees charged separately. OpenAI says GPT-Live-1 gai…
A developer published a tutorial on building a vision-language robot pipeline that combines camera perception with natural-language instructions on NVIDIA Jetson hardware running ROS 2. The guide walks through JetPack an…
Search Engine Journal will host a webinar on Thursday, September 24 at 11 am ET with Google and Uberall on preparing local SEO strategies for Google's upcoming AI search updates, citing that 68% of brands are missing fro…
Java software engineer Adam Rozycki, based in France with an academic background in philosophy and mathematics, posted to the Hugging Face forums seeking a collaborator with LLM architecture experience to build novel LLM…
OpenAI released GPT-Live-1 as a developer API, a full-duplex speech model that listens and talks simultaneously and is already running inside ChatGPT, priced at $0.05 per minute. On OpenAI's benchmarks, GPT-Live-1 scores…
Researchers introduced PARSER, a method that decouples document traversal from reasoning depth to address the sensitivity to evidence placement and linear inference latency that affect sequential memory agents on long do…
Garrett Jones published a prompt, "Language Root Translator," on Hacker News' Show HN that turns any sufficiently strong chatbot into a translation service with language-learning features. The prompt requires users to sp…
Nari Labs launched a Free Public Beta of realtime text-to-speech endpoints built on its own inference engine for the Qwen3-TTS model, claiming 50 ms time to first audio in Fast mode versus 300 ms for ElevenLabs TTS v3 an…
A technical breakdown of how tokenizer training data skewed toward English — with English and code making up 95% of Llama 3's training data — creates measurable cost, context-window, and accuracy penalties for other lang…
The Georgian AI Lab benchmarked six agentic search tool configurations and found that embedding re-ranking inside the search tool lifts evidence recall by roughly 27% versus the same tool without one, with the best setup…
ElevenLabs released Dubbing v2, a new AI dubbing model that conditions directly on original audio performances rather than text transcripts, preserving tone, pacing, and emotion across more than 90 languages. The model h…
A developer built a supervised machine learning pipeline that classifies Apple-related tweets as negative, neutral, or positive using the Apple Twitter Sentiment dataset. The system cleans raw tweet text, converts it int…
BuildBetter, a customer-conversation analysis company founded by Spencer, has spent roughly $7M over six years building a system to read every customer conversation across an enterprise, according to a blueprint the foun…
A developer has mapped Jakob Nielsen's classic UX response-time thresholds onto voice AI, arguing that the 100ms and 1-second cliffs contract sharply in voice interfaces because silence carries no progress signal, with a…
A paper by Federico Barbero and co-authors, revised 13 May 2025 as v3 on arXiv, argues that Rotary Positional Encodings (RoPE) in Transformer-based large language models are not primarily useful for decaying token depend…
A developer argues that computing is shifting from an application-centric model, where users open individual apps on a smartphone, toward an agent-centric model in which AI agents interpret user intent and orchestrate to…
Google changed its Maps content policy in April, barring merchants from asking staff to request a set number of reviews or reviews that mention specific details such as a staff member's name, Moz VP of Revenue Jonathan B…
Non-invasive speech decoding remains constrained by the low signal-to-noise ratio of neural recordings, which makes fine-grained reconstruction of phonemes or individual words difficult, according to research motivated b…
A new method called AgentGrad applies intervention-guided prompt optimization to large language model-based multi-agent systems, according to the paper's abstract. The approach targets the prompt design of each specializ…