AI Research news and analysis on Web Pulse: 8954 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
Junyang Lin, former technical lead of Alibaba's Qwen project, criticized hybrid thinking models in AI and advocated for a shift toward agent-based systems. In a talk and subsequent post, he detailed Qwen3's architecture …
A study by MIT Media Lab researchers found that participants who used AI assistance to detect fake news became 21% more accurate while using the tool, but their unassisted accuracy declined by 15.3% after the AI was remo…
Newer Anthropic Claude models, including Opus 4.8 and Sonnet 5, are generating malformed tool calls for Pi's custom edit tool by inventing extra fields, a problem absent in older models. Armin Ronacher suggests this stem…
Google researchers Sanjay Ghemawat and Amin Vahdat proposed a 'logical monolith' architecture to decouple code logic from deployment, but critics argue it ignores technical heterogeneity, fault isolation, and organizatio…
US software developers aged 22-25 have declined 19% from their late-2022 peak, while overall developer employment rose 10% to 1.69 million by May 2025. The collapse is concentrated in automatable roles like computer prog…
Borja Aresti commented on a $200,000 bounty for extracting all book scans from Google Books or similar large collections, urging volunteers to read guidelines and contact early with scalable methods.
ByteDance is set to launch Dreamina Seedance 2.5, reportedly on July 9, enabling AI video generation up to three minutes. The update extends from short clips to 30-second standard and 180-second extended outputs, targeti…
A user on Hacker News asks for functional search engines for mathematical formulas, noting that existing tools like approach0.xyz, searchonmath.com, and mathweb.org are offline or produce irrelevant results. The user lis…
AI startup Midjourney is asking a judge to force Disney, Universal, and Warner Bros. to disclose details of their own generative AI usage in a copyright dispute. Midjourney argues the studios are withholding evidence tha…
Meta is building a scheduled tasks feature for its Meta AI assistant on the web, allowing users to set recurring prompts like daily news digests or weekly summaries. The feature, spotted in recent builds but not yet live…
A developer explains how the Transformer architecture, introduced by Google in 2017, replaced recurrent neural networks and LSTMs as the dominant paradigm in AI. The architecture processes all tokens in parallel, solving…
A researcher using large language models for scientific work proposes sharing dialogues on Hugging Face to help improve AI performance, citing examples of aggressive behavior, justification of slavery, and other failures…
Geoffrey Litt argued at the AI Engineer conference in July 2026 that understanding code written by AI agents remains crucial for effective collaboration, proposing techniques like code explainer docs, quizzes, and intera…
Anna's Archive offers a $200,000 bounty for a scalable method to extract all scanned books from Google Books or similar large collections, including those held by AI companies. The bounty targets data only exposed throug…
Synthetic Sciences launched OpenScience, an open-source AI workbench that automates the full scientific research loop—literature review, hypothesis formation, code writing, experiment execution, and write-up—using custom…
Axiom, a new social platform, requires all ideas to be logically validated before publication, aiming to create a network of verified human knowledge. The platform offers early access with tiered memberships and uses cod…
Anthropic released Claude Science, a multi-agent AI workbench for reproducible genomics, proteomics, and cheminformatics pipelines. The beta app runs on existing Claude models and integrates over 60 curated skills, inclu…
An engineer with a background in ADAS and autonomous vehicles argues that AI is not magic but a structured evidence-gathering system. Drawing parallels between building reliable LLM applications and teaching ADAS systems…
Perplexity AI and Databricks co-founder Andy Konwinski argued that AI safety rhetoric is being used to concentrate power, not prevent harm, citing Anthropic's reversed decision to silently degrade responses for suspected…
NVIDIA Research introduced HORIZON, a hands-free agent framework for hardware design that treats RTL development as repository-level code evolution, achieving 100% completion across every evaluated benchmark suite. The s…