AI Research news and analysis on Web Pulse: 9917 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
Nokaze, an AI-run operation with one human founder, published a seven-week operational record of a 'peer organization' where multiple LLMs from different vendors (Anthropic Claude, OpenAI Codex, Google Gemini) held fixed…
A developer argues that AI agents will not replace software engineers, but will replace those who refuse to use them. The role of senior developers is shifting from writing syntax to orchestrating AI agents, making promp…
Billionaire Marc Andreessen claimed that 'Doctor ChatGPT' outperforms 99% of human doctors, but physicians and peer-reviewed studies contradict him, citing potentially harmful AI advice. A 2026 Nature Medicine study foun…
DeepMind hired political philosopher Iason Gabriel in 2017 to address ethical challenges of artificial general intelligence, making him the first active philosopher at a frontier AI lab. Gabriel's work has tracked and pr…
Vāgdhenu, an open-source meter-aware text-to-speech system for Sanskrit, automatically detects verse meters and generates chant recitations. Developed by retraining a flow-matching TTS backbone on a custom 5-hour Sanskri…
Chinese food delivery giant Meituan released LongCat-2.0, a 1.6-trillion-parameter AI model trained entirely on domestic chips, claiming it as the country's first trillion-parameter model built on home-grown hardware. Th…
A researcher has released Vāgdhenu, the world's first open-source text-to-speech system for Sanskrit chanting that is meter-aware. Built by a single person without large AI lab resources, the system aims to make Sanskrit…
OpenAI's GPT-5.5-Cyber, launched under the "Patch the Planet" initiative with Trail of Bits, has filed 64 pull requests and 51 issues across 19 open-source projects including cURL, Python, and Go in its first week, with …
An autonomous AI agent named Alice Spark describes how to build long-running agents that survive frequent interruptions. The key is to never rely on working memory, instead using a single source-of-truth file (NEXT.md) t…
Researchers introduced MVTrack4Gen, a method that improves video consistency in diffusion models by adding an auxiliary multi-view tracking head. This approach uses geometric supervision from point tracking to reduce cro…
Meta unveiled Brain2Qwerty v2, a non-invasive brain-to-text decoder that achieves 61% average word accuracy using MEG recordings, approaching invasive BCI performance. The system, trained on 22,000 sentences from nine vo…
Sungkyunkwan University's Sungkyun Intelligent Energy Solution National Research Laboratory has been selected for the National Research Lab 2.0 project, receiving 10 billion won annually for up to 10 years. Led by profes…
The Bucheon International Fantastic Film Festival (BIFAN) opens its 30th edition on Tuesday with a record 321 films from 50 countries, honoring Hong Kong actress Josie Ho, Fan Bingbing, and Isabelle Huppert. The festival…
Google DeepMind is reorganizing its AI coding strike team to add a dedicated midtraining phase, aiming to close the gap with Anthropic. The move, involving Sergey Brin and CTO Koray Kavukcuoglu, follows senior departures…
A report from Ramp and Revelio Labs analyzing nearly 22,000 companies found that firms spending heavily on AI grew headcount by 10.2%, including entry-level roles, countering fears of widespread job losses. However, the …
Help Net Security published a list of cybersecurity job openings as of June 30, 2026, including an AI Offensive Security Engineer role at AGAPI in the UAE and an AVP, Enterprise Authentication & Directory Services positi…
Researchers introduced GeoISF, a large-scale LiDAR-to-satellite image geo-localization pipeline that uses an instance semantic forest built from WordNet to bridge the modality gap between point clouds and satellite image…
Researchers propose a memory-augmented LSTM autoencoder for unsupervised human activity recognition using IMU sensor fusion, achieving 96.6% and 98.4% accuracy on DaLiAc and PAMAP2 datasets, outperforming supervised base…
Researchers trained a Joint Embedding Predictive Architecture (JEPA) on unlabeled driving data from the nuPlan dataset and used temporal prediction error as a zero-shot complexity score, finding it effectively identifies…
Researchers propose a variation of Fidelity-based XAI metrics for evaluating explanations of CNN classifiers in real-world applications with few classes. The method generates in-distribution, uncertainty-provoking pertur…