Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.
OpenAI's Prompts feature within team workspaces for ChatGPT appears to be aimed at organizations. It creates a library of shared prompts made by team members accessible across the team. Team members are able to create prompt templates, set custom instructions, and integrate variables. Teams will be able to package up complex, multistep tasks into single prompts using built-in tools.
Anthropic's new Skills feature consumes only dozens of tokens per skill until needed compared to MCPs that can burn tens of thousands upfront. The design is incredibly simple: drop instructions in a folder and Claude figures out when to use them. This token efficiency and model-agnostic approach (any coding agent can read the files) will trigger a "Cambrian explosion" that makes the MCP rush look pedestrian, especially since skills transform Claude Code from a coding tool into a general automation agent capable of handling anything you could accomplish by typing commands.
GPT-5-level models now match human professionals on roughly half of economically valuable tasks, completing work in minutes that would take a human hours. OpenAI has proposed a retraining program targeting 10 million Americans by 2030, acknowledging evidence that AI is already displacing entry-level workers. The company frames AI as "more enabler than replacer" despite proposing a massive retraining initiative, which has historically failed to prevent job displacement from technological change.
Sponsored
Cognition trained specialized models that retrieve code context 10x faster than frontier models by using RL to issue up to 8 parallel tool calls per turn instead of the sequential searches typical agents use. Speed matters more than the industry realizes because it allows developers to stay in a flow state.
ScaleRL is a framework for analyzing and predicting compute scaling in LLM reinforcement learning. Their large-scale study shows that stable training recipes yield predictable performance curves, aiding efficient resource allocation.
Recursive Language Models (RLMs) can decompose and recursively interact with their input context as a variable. An RLM using GPT-5-mini outperforms GPT-5 with more than double the number of correct answers on OOLONG while being cheaper per query on average. RLMs outperform other methods ReAct + test-time indexing and retrieval over the prompt without any degradation in performance when given over 10 million tokens at inference time. RLMs trained explicitly to recursively reason are likely to be the next milestone in general-purpose inference time scaling.
Manus 1.5 adds full-stack web development that handles backend infrastructure, databases, QA, and user authentication entirely, achieved entirely through natural language. The agent can test apps by itself and fix issues autonomously.
Sponsored
100+ million groups of images, videos, text, audio, and 3D point
DeepSomatic is a convolutional neural network tool that identifies somatic variants in tumor cells more accurately across multiple sequencing platforms.
Claude now connects with Microsoft 365, enabling search across SharePoint, Outlook, and Teams to surface insights and streamline decision-making within enterprise workflows.
Meta's Backbone network provides wide-area network connectivity to networked locations.
The US is estimated to deliver almost 80 gigawatts of new power generation starting in 2027.
US Uber drivers and couriers can now earn extra money by performing 'microtasks' to train AI models.
Character.AI has released Scenes, a structured feature that lets users create interactive, narrative-driven role-play settings without needing to build a new character each time.
Sponsored
You want AI to become part of your organizational DNA - and that means enabling every department to build out their own use cases, without IT gatekeepers standing in the way. But it shouldn't mean an ungoverned free-for-all.
Anthropic's new small model delivers Claude Sonnet 4 performance at one-third the cost and twice the speed. Haiku 4.5 enables new multi-agent workflows where Sonnet 4.5 orchestrates teams of Haiku models working in parallel.
Google has introduced Veo 3.1 with improved audio output, better prompt adherence, and more realistic video generation. The update enhances Google's Flow editor with new capabilities like object-level control, image-to-video blending, and audio-enabled video extensions.
Nvidia, Microsoft, BlackRock, and xAI will acquire Aligned Data Centers for $40 billion, marking the largest global data center deal. The acquisition is AIP's first investment and aims to support AI infrastructure expansion. The deal is expected to close late next year, pending regulatory approvals.
General methods that leverage computation are ultimately the most effective by a large margin. Developers building and working with AI applications should be aware that the artisanal architecture they are building may become obsolete with the next model release. They should make a note of what is not scaling with compute and focus their artisanal ingenuity there. Custom architectures may be necessary for many applications, but build them with the knowledge that they may have to be thrown away when the next frontier model drops.
The software that manages the mathematical definitions in CAD software is called a geometry kernel. Whoever first develops a truly capable foundational AI model for CAD will control a critical layer of industry infrastructure. The problem is that this critical infrastructure layer is currently monopolized by a handful of players whose business depends on maintaining closed, proprietary systems. If the US wants to maintain its industrial competitiveness, it must recognize geometry kernels as strategic infrastructure, not just commercial software products. The market has had decades to produce an open alternative and hasn't, so a coordinated public investment may be justified.
Anthropic gathered policy ideas from economists and experts to prepare for modest to dramatic job losses due to AI. Concrete proposals include $10,000 annual employer training grants, "token taxes" on AI-generated content, and sovereign wealth funds that give citizens stakes in AI revenues.
Analysis of 65,000 web articles found AI-generated content surpassed human-written articles in November 2024, growing from near-zero to 39% within 12 months of ChatGPT's launch before plateauing since May 2024. The authors speculate the stagnation was due to poor SEO performance, but they were unable to measure the prevalence of AI-generated, human-edited articles.
Sponsored
Access 500+ AI models from with a single API + SDK. Swap LLMs on a prompt-by-prompt basis [12] to control costs, improve quality, and guarantee uptime. Build without vendor lock-in and without managing
Official Python SDK for Dedalus Labs' AI agent platform, providing APIs and tools to build, manage, and interact with autonomous agents and workflows.
AI models collapse into repetitive, generic responses because human annotators systematically prefer familiar answers during training. This "typicality bias" explains why ChatGPT, Claude, and Gemini all tell the same coffee joke. The solution is simple: prompt models to "generate 5 jokes with their corresponding probabilities." This "verbalized sampling" method increased creative diversity by roughly 2x and improved human evaluation scores by 26% across tasks from story writing to social simulation.
Oracle plans to offer computing firepower and a suite of bundled services that will make artificial intelligence useful for business. Investors have expressed concerns about how its newest deals rely heavily on OpenAI, which won't generate a profit until 2029. Oracle's long-term play is to sell an ecosystem of AI solutions to large enterprises. The company has taken on debt to finance its AI infrastructure build-out, further troubling investors. However, the company has a more varied business than many realize - pretty much all of the big model providers use its cloud in one form or another.
Starcloud, under NVIDIA Inception, plans to launch AI-equipped satellites to establish data centers in space, aiming for 10x lower energy costs and higher sustainability. Its 5-gigawatt orbital center will leverage solar energy and space's vacuum for cooling, conserving Earth's resources. This approach offers faster processing for applications like wildfire detection and Earth observation, utilizing NVIDIA's advanced GPUs for real-time data inference.
Meta is migrating its AI ranking and recommendation systems to Arm's Neoverse platform as part of a multi-year partnership aimed at scaling AI infrastructure. The deal focuses on energy-efficient cloud AI deployments, without involving ownership stakes or shared physical assets.
Sponsored
You'll probably like our flagship newsletter. It's all about tech, science, and programming.
The formal request stated that characters from manga and anime are 'irreplaceable treasures' that Japan boasts to the world.
Ke Yang, who headed Apple's Answers, Knowledge, and Information team, has left to join Meta.
Anthropic is aiming to nearly triple its revenue by 2026.
David Sacks believes that Anthropic is positioning itself as a pro-regulation company to push policies that it would benefit from while stifling others who want to get started in the AI sector.
Sponsored
Today, Baseten is releasing GLM 4.6 on Model APIs. Baseten Model APIs are purpose-built for production workloads with Baseten's Inference Stack. Baseten powers the likes of Zed, Sourcegraph and more with leading latency and throughput with proprietary speculative decoding techniques. Unlock out-of-the-box performance with Model APIs:
Walmart has reached a deal with OpenAI to allow shoppers to buy items directly through ChatGPT. The US grocer is trying to keep up with changes in how consumers discover items to buy. OpenAI's Instant Checkout feature, announced in September, has been a new moneymaker for the company. OpenAI charges companies a fee for transactions completed through ChatGPT.
AMD is partnering with Oracle to create a large data-center cluster with the equivalent of 200 megawatts of computing power. AMD plans to deploy 50,000 of its MI450 chips inside data centers run by Oracle. The companies plan to expand their partnership in 2027 and beyond.
Broadcom's deal to develop huge numbers of chips and computing systems with OpenAI is not a clear-cut win. OpenAI has recently signed multiple multi-billion-dollar deals to create data centers stocked with hundreds of thousands of chips. While the scale is remarkable, it is unclear how OpenAI will pay for all of it. OpenAI's revenue this year is expected to be around $13 billion, a large sum, but nowhere close enough to justify this level of spending. The company has told investors it won't be profitable until 2019.
Newer models are becoming more adept at manipulating text at the character level, despite their understanding of text being based on tokens. Most current-generation models are able to decode Base64 text. They are now able to substitute characters fairly reliably. Character-level operations are far from solved, but it is clear they have made huge progress in the area.
The public thinks AI bugs work like software bugs: find the missing semicolon, fix the problem, move on. However, AI failures emerge from training on 11.25 trillion words that would take 85,000 years to read, not buggy code lines. Unlike traditional software, where you can pinpoint and permanently fix problems, AI mistakes can't be traced, "fixed" behaviors can resurface with different prompts, and nobody, including the labs, understands why these systems make the errors they do. Your boss assumes AI problems are patchable because 40 years of software education taught everyone that bugs have discoverable causes and permanent solutions.
Latent diffusion models get an upgrade with Representation Autoencoders (RAEs), which swap out the standard VAE for pretrained encoders like DINO, SigLIP, or MAE paired with learned decoders. These higher-dimensional latent spaces improve generation quality, and new methods allow Diffusion Transformers to handle them effectively.
Sponsored
The next generation of AI won't just respond—it will anticipate, strategize, and act autonomously. Amazon Science analyzes the scientific foundations driving this transformation, offering insights into how agentic systems will reshape how we interact with technology in commerce, logistics, and beyond.