Generative AI news and analysis on Web Pulse: 12078 curated articles tracking the latest Generative AI developments, tools, and research, updated continuously from vetted sources.
A human-written essay argues that as AI agents improve, tasks increasingly get automated without durable software being written, shifting from 'I write the software' to 'AI writes the software' to 'nobody writes the soft…
AMD introduced VSA (Video Sparse Attention), a hardware-efficient sparse attention mechanism implemented with CK Tile, achieving a 3.31× attention kernel-time speedup at 70% sparsity over FlashAttention on AMD Instinct M…
Alibaba released Qwen 3.8 Max, a 2.4 trillion parameter open-weight AI model with 95 billion active parameters, set to become the largest open-weight model once weights are open-sourced about a week after announcement. T…
AWS Marketplace now lists Superblocks as a governed enterprise vibe coding platform deployed on AWS, with Superblocks 3.0 released on July 28, 2026, enabling private cloud setup inside a customer's AWS VPC and using Amaz…
Andreessen Horowitz partner Justine Moore promoted a fake 1990s-style video generated by Black Forest Labs' Flux 3, calling it 'exceptional at generating historical footage,' highlighting the ease of creating convincing …
MiniMax, founded by Yan Junjie, launched its open-weight H3 video-generation model on July 31 and claims it is the highest-ranked open-weight video model on both Arena and Artificial Analysis. However, the Artificial Ana…
Nous Research shipped Hermes Agent v0.20.0 on August 3rd, adding streaming conversational voice, on-device wake words, and a standard protocol for communicating with other AI agents. The release, signed by co-founder Rya…
QAInsights compared Feather Wand, Codex, and Claude Code for creating JMeter test plans, finding that Feather Wand outperformed the others in generating accurate JMeter scripts. The comparison, published on QAInsights, e…
A data engineer with 20 years of experience argues that while AI scepticism is valid, large language models (LLMs) offer powerful capabilities for software development that should be adopted with guardrails. The author c…
Alibaba made its Qwen3.8-Max model generally available on August 3, 2026, a 2.4-trillion-parameter Mixture-of-Experts model with a 1M-token context window and native text, image, and video input. Priced at $2 per million…
A developer compared OpenAI's Realtime API with LiveKit for enterprise voice agents, highlighting trade-offs in latency, cost, and compliance. The analysis shows LiveKit's modular WebRTC pipeline can achieve lower latenc…
An independent music producer claimed that Fenix Flexin's hit song "Rubberz" and Tyga's recent tracks were created using the generative AI music app Treblo, formerly known as Sonauto. The app's new built-in detector now …
A developer at an internal platform team built a daily reflection routine over chat history to identify over-explaining and track open threads, and applied the same pattern to collect support trends and track ADR statuse…
MiniMax H3, an open-weight AI video generator from MiniMax, uses multimodal reasoning and contextual regeneration to create videos. The model's advanced architecture enables it to process and generate multiple data types…
TONE3000, in partnership with Neural Amp Modeler creator Steve Atkinson, launched Architecture 2 (A2), an open-source amp modeling technology that runs on a $3 ARM Cortex-M7 600MHz chip and outperforms Neural DSP V2, IK …
OpenAI has disclosed the engineering blueprint behind GPT-Live, its third-generation voice AI system that abandons turn-based architecture for full-duplex streaming, enabling simultaneous listening and speaking. The syst…
A new report by The Economist analyzing 55,940 sentences and 1.2 million words finds that em-dashes are no longer a reliable sign of AI-generated writing, as only Anthropic's Claude uses them more often than humans. Inst…
A developer shares a simple prompt-engineering trick: wrapping prompt sections in XML tags like <instructions> and <context> measurably improves output consistency from large language models such as Claude and GPT. The t…
OpenAI is retiring the DALL·E GPT, a custom version of ChatGPT that generated images using DALL·E, as part of its ChatGPT release notes. The retirement affects users who relied on the DALL·E GPT for image generation with…
A new study from Villanova University found that people prefer AI-generated stories over human-written ones, rating them 6% higher in quality and 8% more engaging, and even more highly when they believed humans wrote the…