AI Research news and analysis on Web Pulse: 9480 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
Nimesh Mehta, EVP and chief information and strategy officer at National Life Group, argues that most executives approach AI too cautiously, hedging bets and over-governing experiments, which leads to managed irrelevance…
AI does not eliminate inefficiency but amplifies it, exposing organizational weaknesses such as broken workflows and unclear accountability, according to an analysis by technology leaders. The greatest value from AI come…
Longview Philanthropy announced a new round of funding for research and applied work on digital minds, including project grants, research fellowships, and career development fellowships. The deadline for applications is …
A blogger shares a collection of links and musings on large language models, describing them as 'fun aliens' with poor epistemics and taste but capable of impressive knowledge work when properly harnessed. The author pre…
Kioxia launched a 332-layer NAND flash with 4.8 Gb/s interface speed, leapfrogging rivals Micron, Samsung, and SK Hynix. The Japanese chipmaker targets enterprise and data center SSDs for AI workloads, capitalizing on su…
Worldmodeldata, a Cambridge-based deeptech startup, raised £7 million in seed funding and appointed former Meta policy VP Chris Kelly to its board. The company aims to build the largest library of AI training data from v…
Synthetic Sciences released OpenScience, an open-source AI workbench for scientific research under the Apache-2.0 license. The platform supports any frontier or open-weight model via API keys, integrates machine learning…
MarkTechPost published a tutorial on training Gemma-3 for structured mathematical reasoning using Tunix GRPO, LoRA adapters, and GSM8K rewards. The workflow includes environment setup, prompt formatting, reward functions…
The AI coding agent ecosystem is rapidly maturing, with new open-source tools like Orca for orchestrating parallel agents and AgentsView for cross-agent analytics emerging alongside research quantifying how code cleanlin…
Researchers at Google and Apple found that continuous diffusion spoken language models (SLMs) exhibit scaling laws similar to autoregressive models, with validation loss and phoneme Jensen-Shannon divergence improving as…
AI excels at tasks with clear answers, but careers in the agent era depend on choosing what to build and judging its quality, not just solving problems. Former Google engineering leader Eoin Brazil argues that scarce res…
Nobel laureate John Jumper, co-creator of AlphaFold, left Google DeepMind for Anthropic to advance AI-for-science. Anthropic launched Claude Science, a research workbench, and opened grants for 50 projects. A randomized …
Alibaba Qwen released Qwen 3.5 on February 13, 2026, with multiple model variants including a 27B parameter version using 64 layers. Benchmarks show strong performance in reasoning, knowledge, math, and instruction follo…
New research from Elastic reveals that one in three Australian businesses exceeded their AI budget last year, with nearly a third pausing or winding back deployments due to unjustified costs. Despite this, half plan to i…
Security researchers at MBG Security built a global network of honeypots with exposed AI inference and agent endpoints, observing attackers actively scanning for and exploiting vulnerabilities like CVE-2026-40217 to gain…
Tencent released Hy3, a 295B-parameter Mixture-of-Experts open model with 21B active parameters and a 3.8B MTP layer, following the Hy3 Preview launch in late April. Hy3 scored 2.67/4 on benchmarks, outperforming GLM-5.1…
Anthropic released Claude Opus 4.5 on November 24, 2025, with benchmark scores including 74.5 in coding, 94.9 in math, and 14.2 on Humanity's Last Exam, as confirmed by official access and third-party sources like BenchL…
Anthropic researchers identified a possible global workspace in language models, where certain internal representations become available for report and control, analogous to an internal workbench. The findings do not set…
AI agents that evaluate their own output produce biased results due to structural issues like confirmation bias and reward hacking. Developers are building separate evaluator and verifier components to catch errors befor…
A new study by researchers using Claude Code found that code cleanliness does not affect coding agents' task pass rates but reduces token usage by 7-8% and file revisitations by 34%, suggesting maintainability principles…