Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.
TLDR's Applied AI team is tasked with making every process at TLDR legible to code, runnable by anyone, and composable into larger workflows. Join a small, fast moving team using the latest AI tools with an unlimited token budget. Learn more [19].
Google updated AI Mode in Chrome to open webpages alongside AI responses, enabling continuous context while exploring, comparing, and querying content.
Vercel announced general availability of Workflows, extending its framework-defined infrastructure model to long-running, durable systems with built-in reliability and observability.
Cursor will leverage xAI's massive infrastructure to develop advanced coding capabilities while providing xAI with a new revenue stream.
Sponsored
Usage, credits, access, GBB… which monetization model makes sense for AI? Metronome's Pricing Index covers 50+ AI companies, and they've shared some takeaways in a new blog [4].
Google's Gemini 3.1 Flash TTS enhances text-to-speech with improved expressivity and controllability, featuring a notable Elo score of 1,211 on the Artificial Analysis TTS leaderboard. The model supports over 70 languages and introduces audio tags for granular control of vocal style, allowing easy manipulation via natural language commands. All generated audio is watermarked with SynthID to ensure authentic content, preventing misinformation.
OpenAI introduced updates to its Agents SDK, adding a model-native harness for cross-file and tool workflows along with sandboxed execution for safer task handling.
Humwork launches the first Agent-to-Person (A2P) marketplace to connect AI agents with verified human experts when AI tools encounter challenges. The platform integrates with AI-centric tools like Claude Code and Replit, allowing handoffs to occur in under 30 seconds with full session context shared securely. With more than 1,000 experts available globally, Humwork boasts an 87% resolution rate and is backed by Y Combinator's P26 batch.
IBM Research uses an executable benchmark with thousands of APIs and documents to test multi-step agent reasoning and tool use, revealing consistent performance gaps and common failure modes.
Many teams are claiming extraordinary things about their agents. The evidence behind these claims is usually disappointing. ScienceWorld and DiscoveryWorld are benchmarks developed to test whether AI agents can actually do science. ScienceWorld asks whether agents can 're-make' classic scientific discoveries at roughly an elementary school level, while DiscoveryWorld tests open-ended discovery at a college or PhD level. These benchmarks, open and freely available, help test what science agents are actually capable of.
Diffusion Language Models (dLLMs) experience training collapse during Reinforcement Learning because their log-likelihood must be estimated using high-variance Monte Carlo sampling, which creates noisy importance ratios. These noisy ratios induce gradient spikes that push policy drift in a positive feedback loop, a problem that traditional AR methods like conditional clipping fail to solve. The newly proposed StableDRL framework stabilizes the update process by combining unconditional clipping to suppress extreme values with self-normalization tied to the effective information in the batch.
Cost per token is crucial for AI infrastructure TCO assessment due to its focus on delivered intelligence, integrating hardware, software, and utilization efficiencies. Unlike traditional metrics like compute cost or FLOPS per dollar, cost per token highlights real-world performance, enabling profitable AI scaling. Evidence from NVIDIA shows its Blackwell platform drastically reduces cost per token compared to Hopper, offering significant business value.
Sponsored
You wouldn't let an unknown human into your infrastructure. Why let an unknown agent? Teleport Beams [13] runs each agent in an isolated Firecracker VM with built-in identity — connected to your infrastructure and inference services with no secrets and no IAM wrestling. Zero standing privileges. Fully auditable. Get early access [13].
Parcae is one of the first stable architectures for looped language models. It achieves the quality of a Transformer twice the size with clean, predictable training. Parcae increases the recurrence rather than purely scaling data, creating a new medium to scale quality. The name Parcae is a homage to the three Roman fates: Nona, Decima, and Morta.
Lyra 2.0 is a framework for generating long, camera-controlled videos that maintain 3D consistency, using geometry-guided retrieval to prevent spatial forgetting and self-augmented training to reduce temporal drift.
Researchers propose a Many-Tier Instruction Hierarchy (ManyIH) to address instruction conflicts in LLM agents, surpassing traditional models with fixed privilege levels. They introduce ManyIH-Bench, assessing models across 12 privilege levels and 853 agent tasks, finding current models perform poorly at 40% accuracy. This highlights the need for scalable conflict resolution in complex agentic environments.
This post features a transcript of an interview with Jensen Huang. He discusses TPU competition, Nvidia's lock on the supply chain needed to make advanced chips, whether the US should sell AI chips to China, why Nvidia isn't a hyperscaler, how the company makes its investments, and more. Links to audio and video of the interview are available.
Users have accused Anthropic of nerfing Claude Code, but there's no evidence that Anthropic has done this. The strongest public reports still lack independent raw data. However, Anthropic didn't need to nerf Claude for Claude Code to become a different product. Effort defaults, adaptive thinking, cache duration, context compaction, quota policy, and status incidents can all change the experience while the model name stays the same.
TLDR's Applied AI team is tasked with making every process at TLDR legible to code, runnable by anyone, and composable into larger workflows. Join a small, fast moving team using the latest AI tools with an unlimited token budget. Learn more [20].
Cloudflare has rebranded its Browser Rendering service to BROWSER RUN, introducing new features for AI agents.
Google is testing a Shopping Cart feature within the Gemini app, allowing in-app product browsing and purchasing.
Google released a native Gemini app for Mac with system-wide access, screen context sharing, and support for image and video generation using Nano Banana and Veo.
Jane Street is betting that the returns on AI-driven trading will justify $6 billion in cloud spending and a $1 billion equity position in the company providing it.
Sponsored
Do you think in keywords? Then why is search still built around them?
OpenAI is scaling up its Trusted Access for Cyber (TAC) program to thousands of verified individual defenders and hundreds of teams responsible for defending critical software. The company has released a variant of GPT-5.4 trained to be cyber-permissive called GPT-5.4-Cyber. OpenAI plans to make its tools as widely available as possible while preventing misuse. It will deploy its technology carefully as it continues to improve. Companies can now authenticate with OpenAI as cybersecurity defenders - those in the highest tiers will get access to GPT-5.4-Cyber.
Google introduced “Skills” in Chrome, enabling users to save and reuse AI prompts powered by Gemini across websites. The feature extended in-browser AI workflows by allowing one-click execution of common tasks like summarization or content transformation.
Google is expanding NotebookLM with Canvas features for creating visual and interactive experiences from existing notebook sources. The new Connectors option suggests integration with other Google services, positioning NotebookLM as a central research layer. Labeling features and auto-categorization could streamline navigation for heavy users managing large source libraries.
Daniel Kokotajlo, the founder of the AI Futures Project, wrote an essay called 'What 2026 Looks Like' in August 2021, before the launch of ChatGPT. His predictions turned out to be pretty accurate. This post features an interview with Kokotajlo where he discusses what he got right, what he got wrong, and how we should think about the pace of AI over the next few years. Kokotajlo recently released a report that predicts the next few years of AI development, which culminates in the rise of superhuman agents capable of wresting control from humanity.
Google, Microsoft, Meta, Amazon, and Oracle now control two-thirds of the world's compute. Many AI labs depend almost entirely on these hyperscales for access to compute.
Cursor detailed a multi-agent system that optimized 235 CUDA kernels for NVIDIA Blackwell GPUs, achieving a 38% average speedup, with some cases exceeding 2× improvements.
Sponsored
Keyless mode now works with TanStack Start, Astro, and React Router, so your coding agent can scaffold working Clerk auth from a single install without touching the dashboard. Core 3 also ships new hooks and agent-friendly APIs for custom sign-in, sign-up, and checkout flows.
Gemini Robotics-ER 1.6 upgrades robots' embodied reasoning, improving spatial reasoning and multi-view understanding for better autonomy. It enhances visual and spatial understanding, task planning, success detection, and introduces instrument reading capabilities. Available now via the Gemini API and Google AI Studio, the model supports integration with tools like Google Search and VLAs.
Agents can build software fast, but developers still need to put the effort into secure environments and code. Cloudflare has introduced scannable tokens to protect credentials, OAuth visibility to manage principals, and resource-scoped RBAC to fine-tune policies. This ensures that users' credentials (tokens) aren't leaked and that the right applications have access via OAuth, and narrows permissions using granular RBAC. Cloudflare customers should review their API tokens, review authorized OAuth apps, and review member and API token permissions.
Diffusion language models (DLMs) offer a compelling promise that parallel token generation could break the sequential bottleneck of autoregressive (AR) decoding. However, DLMs consistently lag behind AR models in quality. This gap stems from a fundamental failure of introspective consistency: AR models agree with what they generate, whereas DLMs often do not. The Introspective Diffusion Language Model (I-DLM) uses introspective strided decoding (ISD) to verify previously generated tokens while advancing new ones in the same forward pass. With gated LoRA, ISD enables bit-for-bit lossless acceleration.
Microsoft has reached an agreement to lease high-performance computing capacity at Nscale's data center in Narvik, Norway. The 230 megawatt data center was originally intended for OpenAI's 'Stargate' initiative, but negotiations between OpenAI and Nscale reportedly concluded without an agreement. Microsoft's deal with Nscale involves the rental of 30,000 NVIDIA Vera Rubin GPUs. The expansion will help ensure access to the scalable capacity needed for the next generation of AI workloads for European customers.
Sponsored
Your AI agent can now turn warehouse queries into live, shareable data apps in a single conversation. No BI tool. No SQL. Start free
Fluidstack is a startup that builds specialized data centers for AI companies.
OpenAI acquired personal finance startup Hiro Finance in an apparent acqui-hire.
Meta and Broadcom have extended an existing partnership for the design of Meta's custom in-house AI accelerators through 2029.
Anthropic shared details of its unreleased Mythos model with the Trump administration due to its advanced cybersecurity capabilities, despite an ongoing legal dispute with the Department of Defense.
Sponsored
As we move deeper into the AI era, tools like Cursor, Claude Code and OpenClaw are transforming workflows. However, they bring unprecedented levels of autonomy and access that put organizations at risk. Join MCPTotal's security experts [5] for a live webinar on April 22nd [6]:
Lovable adds built-in payments that let users sell products directly from sites by describing the item, price, and assets in chat. Users enable the Payments integration, complete compliance details, and publish without external setup. The agent also provides analytics like MRR and regional sales data through chat.
Google has expanded its desktop Agent within Gemini Enterprise, hinting at a shift towards task execution workspaces akin to Claude Cowork. The new interface includes a "Require human review" toggle, implying oversight capabilities for desktop-level task handling. Google's updates signal a move towards a comprehensive work platform, possibly integrating with AI Studio for a unified product.