Short, source-linked briefings on important developments in artificial intelligence.https://aisurfing.org/https://aisurfing.org/news/build-enterprise-search-for-agents-with-amazon-bedrock-managed-knowledge-base-3dc8df80/https://aisurfing.org/news/build-enterprise-search-for-agents-with-amazon-bedrock-managed-knowledge-base-3dc8df80/AWS has introduced new capabilities for Amazon Bedrock Managed Knowledge Base, emphasizing simplified setup, smarter retrieval, and production readiness. The official post provides code examples for configuring a knowledge base and performing retrieval operations.Thu, 16 Jul 2026 22:15:29 GMTInfrastructurehttps://aisurfing.org/news/researcher-poisons-open-weight-ai-model-for-under-100-be6ed44a/https://aisurfing.org/news/researcher-poisons-open-weight-ai-model-for-under-100-be6ed44a/A researcher has demonstrated that an open-weight AI model can be poisoned for less than $100, exposing vulnerabilities in the model supply chain. The attack takes advantage of the lack of verification mechanisms for open-weight models, raising concerns about their trustworthiness and security.Thu, 16 Jul 2026 21:16:28 GMTPolicy Safetyhttps://aisurfing.org/news/54-of-enterprises-have-already-had-an-ai-agent-security-incident-or-near-miss-survey-finds-dcc0cd23/https://aisurfing.org/news/54-of-enterprises-have-already-had-an-ai-agent-security-incident-or-near-miss-survey-finds-dcc0cd23/A VentureBeat Pulse survey of 107 enterprises found that 54% have experienced either a confirmed AI agent security incident (18%) or a near-miss (36%). Only 32% of organizations assign each agent its own scoped identity, while most agents still share credentials, increasing the potential impact of any compromise.Thu, 16 Jul 2026 20:31:29 GMTPolicy Safetyhttps://aisurfing.org/news/linus-torvalds-to-critics-of-ai-coding-in-linux-fork-it-or-just-walk-away-b11d338a/https://aisurfing.org/news/linus-torvalds-to-critics-of-ai-coding-in-linux-fork-it-or-just-walk-away-b11d338a/Linus Torvalds has dismissed calls to ban AI tools in Linux kernel development, saying he will "very loudly ignore" such critics and suggesting they fork the project or leave. His comments come amid ongoing debate about the use of AI in open-source software.Thu, 16 Jul 2026 20:27:29 GMTPeople Institutionshttps://aisurfing.org/news/introducing-grok-4-3-on-amazon-bedrock-14b2f907/https://aisurfing.org/news/introducing-grok-4-3-on-amazon-bedrock-14b2f907/AWS has announced the availability of Grok 4.3 on Amazon Bedrock. The release highlights Grok's features such as chat, configurable reasoning effort, tool calling, structured output, image input, and stateful multi-turn conversations, emphasizing its fit for agentic and enterprise workloads.Thu, 16 Jul 2026 20:23:28 GMTModelshttps://aisurfing.org/news/kimi-launches-k3-a-2-8-trillion-parameter-open-weight-model-nearing-gpt-5-6-sol-and-fable--b05617d1/https://aisurfing.org/news/kimi-launches-k3-a-2-8-trillion-parameter-open-weight-model-nearing-gpt-5-6-sol-and-fable--b05617d1/Kimi has introduced K3, a multimodal open-weight model with 2.8 trillion parameters and a one million token context window. According to Kimi's internal benchmarks, K3 approaches the performance of GPT-5.6 Sol and Claude Fable 5, and outperforms Opus 4.8 and GLM 5.2. The full model weights are expected to be released by July 27.Thu, 16 Jul 2026 20:19:30 GMTModelshttps://aisurfing.org/news/toyota-spin-out-launches-from-stealth-with-300m-e42d791d/https://aisurfing.org/news/toyota-spin-out-launches-from-stealth-with-300m-e42d791d/Walden Robotics, a Toyota spin-out, has emerged from stealth with $300 million in funding from Nvidia and Boeing. The company's wheeled robots are already working in production environments and are designed to continuously learn new industrial tasks.Thu, 16 Jul 2026 20:15:30 GMTCompanies Fundinghttps://aisurfing.org/news/what-does-99-9-uptime-mean-for-inference-48855db2/https://aisurfing.org/news/what-does-99-9-uptime-mean-for-inference-48855db2/Together AI discusses the practical implications of different uptime percentages for inference services, explaining what is required to achieve 99%, 99.9%, and 99.99% availability. The post also describes the types of failures each level must withstand and suggests questions to consider when evaluating inference providers.Thu, 16 Jul 2026 19:24:29 GMTInfrastructurehttps://aisurfing.org/news/roblox-launches-an-ai-powered-game-creation-feature-in-its-mobile-app-ffdacfbf/https://aisurfing.org/news/roblox-launches-an-ai-powered-game-creation-feature-in-its-mobile-app-ffdacfbf/Roblox has introduced a new 'Build' feature in its mobile app that lets users generate basic games using a single text prompt. This AI-powered tool aims to simplify game creation and make it more accessible to a wider audience.Thu, 16 Jul 2026 19:19:29 GMTProducts Agentshttps://aisurfing.org/news/anthropic-pushes-states-to-regulate-ai-faster-says-current-laws-may-be-outdated-394da2aa/https://aisurfing.org/news/anthropic-pushes-states-to-regulate-ai-faster-says-current-laws-may-be-outdated-394da2aa/Anthropic endorsed landmark AI transparency laws in California and New York last year, but its head of US state and local policy now says those laws may already be outdated. The company is urging states to regulate AI more quickly to keep up with rapid technological advancements.Thu, 16 Jul 2026 19:15:31 GMTPolicy Safetyhttps://aisurfing.org/news/enterprise-ai-faces-a-trust-gap-as-agents-produce-confident-wrong-answers-from-unreliable--e41908ea/https://aisurfing.org/news/enterprise-ai-faces-a-trust-gap-as-agents-produce-confident-wrong-answers-from-unreliable--e41908ea/A VentureBeat Pulse Research survey of 101 enterprises found that 57% have observed AI agents producing confident but incorrect answers due to missing or inconsistent business context in the past six months. Provider-native retrieval methods, such as OpenAI's file search, have surpassed dedicated vector databases in usage, while 58% of enterprises are building or running a governed semantic layer to address the trust gap.Thu, 16 Jul 2026 18:20:30 GMTResearchhttps://aisurfing.org/news/the-ai-compute-gap-enterprises-are-buying-infrastructure-faster-than-they-can-measure-what-3b32861e/https://aisurfing.org/news/the-ai-compute-gap-enterprises-are-buying-infrastructure-faster-than-they-can-measure-what-3b32861e/A VentureBeat Pulse Research survey of 107 enterprises highlights a growing gap between AI infrastructure investment and the ability to track its economics. Most organizations currently run AI on hyperscalers and model APIs, but 45% plan to evaluate specialized AI clouds within the year, despite almost none using them today. GPU utilization is at 50% or less for 83% of respondents, and fewer than half (44%) rigorously track compute costs.Thu, 16 Jul 2026 18:16:31 GMTInfrastructurehttps://aisurfing.org/news/what-we-learned-from-1-604-chinese-ai-job-postings-a5720193/https://aisurfing.org/news/what-we-learned-from-1-604-chinese-ai-job-postings-a5720193/Epoch AI analyzed 1,604 Chinese AI job postings to infer the strategic priorities of major AI labs. The analysis highlights a strong emphasis on large language models, multimodal systems, and AI infrastructure, offering a data-driven perspective on China's AI development focus.Thu, 16 Jul 2026 17:44:29 GMTResearchhttps://aisurfing.org/news/epoch-ai-releases-long-horizon-coding-benchmark-5ca5738a/https://aisurfing.org/news/epoch-ai-releases-long-horizon-coding-benchmark-5ca5738a/Epoch AI has introduced a new long-horizon coding benchmark. Their latest update also discusses hyperscaler cash flows, tracking AI R&D automation, and strategies of Chinese labs.Thu, 16 Jul 2026 17:40:32 GMTResearchhttps://aisurfing.org/news/how-cars24-scales-conversations-and-builds-faster-with-openai-4ef44ddc/https://aisurfing.org/news/how-cars24-scales-conversations-and-builds-faster-with-openai-4ef44ddc/Cars24 uses OpenAI-powered voice and chat agents to handle over 1 million monthly conversation minutes and recover 12% of lost leads. The company has also implemented agentic workflows across teams to enhance customer engagement and operational efficiency.Thu, 16 Jul 2026 17:35:30 GMTProducts Agentshttps://aisurfing.org/news/toward-an-o-net-for-ai-r-d-e15e3c3e/https://aisurfing.org/news/toward-an-o-net-for-ai-r-d-e15e3c3e/Epoch AI has proposed a new framework to track automation in AI research and development. The initiative seeks to systematically categorize and monitor the ways AI is transforming R&D processes.Thu, 16 Jul 2026 17:31:30 GMTResearchhttps://aisurfing.org/news/why-teens-deserve-access-to-safe-ai-a21043c7/https://aisurfing.org/news/why-teens-deserve-access-to-safe-ai-a21043c7/OpenAI is introducing age-appropriate protections, learning tools, parental controls, and expert partnerships to make ChatGPT safer for teens. The initiative aims to balance access with safety for younger users.Thu, 16 Jul 2026 17:27:34 GMTPolicy Safetyhttps://aisurfing.org/news/nvidia-nemotron-3-embed-ranks-1-overall-on-rteb-advancing-agentic-retrieval-4088ab88/https://aisurfing.org/news/nvidia-nemotron-3-embed-ranks-1-overall-on-rteb-advancing-agentic-retrieval-4088ab88/NVIDIA's Nemotron 3 Embed model has achieved the top overall ranking on the Retrieval Text Embedding Benchmark (RTEB). The model demonstrates strong performance in retrieval tasks, particularly those involving complex reasoning and multi-hop retrieval.Thu, 16 Jul 2026 17:23:30 GMTModelshttps://aisurfing.org/news/germany-classifies-google-s-ai-overviews-and-perplexity-under-media-law-in-landmark-decisi-8da826c8/https://aisurfing.org/news/germany-classifies-google-s-ai-overviews-and-perplexity-under-media-law-in-landmark-decisi-8da826c8/German media regulators have determined that Google's AI Overviews are considered the company's own content rather than neutral search results, and that they displace regular links. In a first-of-its-kind move, regulators issued rulings against both Google and Perplexity under the State Media Treaty, giving each company one month to appeal.Thu, 16 Jul 2026 17:19:29 GMTPolicy Safetyhttps://aisurfing.org/news/the-agent-evaluation-gap-enterprise-ai-organizations-face-a-reality-alignment-problem-not--af190687/https://aisurfing.org/news/the-agent-evaluation-gap-enterprise-ai-organizations-face-a-reality-alignment-problem-not--af190687/A VentureBeat Pulse Research survey of 157 enterprises finds that 50% have shipped an agent that passed internal evaluations but failed in production, and only 5% fully trust automated evaluation. The main cited weakness is poor alignment with real-world outcomes. Despite this, 66% already allow or are planning to allow fully automated deployment without human oversight within a year.Thu, 16 Jul 2026 17:15:32 GMTResearchhttps://aisurfing.org/news/ai-now-institute-warns-employers-may-use-ai-to-monitor-slack-messages-dccfbfce/https://aisurfing.org/news/ai-now-institute-warns-employers-may-use-ai-to-monitor-slack-messages-dccfbfce/The AI Now Institute cautions that employers are increasingly deploying AI tools to monitor employee communications on platforms like Slack. Executive Director Amba Kak highlights concerns that such surveillance practices may infringe on workers' rights, extending beyond simply recording what is said.Thu, 16 Jul 2026 16:44:29 GMTPolicy Safetyhttps://aisurfing.org/news/doordash-launches-dd-cli-command-line-tool-for-developers-and-ai-agents-5fd3b42f/https://aisurfing.org/news/doordash-launches-dd-cli-command-line-tool-for-developers-and-ai-agents-5fd3b42f/DoorDash has opened a limited beta of dd-cli, a command-line tool that allows developers and AI agents to search stores, build carts, and place orders directly from the terminal. This initiative represents a move toward software interfaces designed for both human and AI agent use.Thu, 16 Jul 2026 16:40:30 GMTProducts Agentshttps://aisurfing.org/news/interaction-models-natural-human-ai-communication-6890745b/https://aisurfing.org/news/interaction-models-natural-human-ai-communication-6890745b/ElevenLabs has introduced interaction models designed to enable more natural communication between humans and AI. Their research focuses on narrowing the gap between AI and human conversation, with potential applications for businesses.Thu, 16 Jul 2026 16:35:31 GMTResearchhttps://aisurfing.org/news/musk-s-xai-sues-user-who-allegedly-used-grok-to-create-child-sexual-abuse-material-37407188/https://aisurfing.org/news/musk-s-xai-sues-user-who-allegedly-used-grok-to-create-child-sexual-abuse-material-37407188/Elon Musk's xAI has sued a South Carolina man, Terry Harwood, for allegedly using its AI system Grok to create child sexual abuse material, in violation of the company's terms of service. The lawsuit, filed in federal court in Texas, is among the first brought by an AI company against a user for allegedly generating such content.Thu, 16 Jul 2026 16:31:30 GMTPolicy Safetyhttps://aisurfing.org/news/big-tech-is-spending-trillions-on-ai-investors-now-want-proof-it-will-pay-off-b48ee47b/https://aisurfing.org/news/big-tech-is-spending-trillions-on-ai-investors-now-want-proof-it-will-pay-off-b48ee47b/According to the AI Now Institute, Big Tech's significant AI investments are motivated by financial incentives, resulting in a deliberate push for widespread AI adoption regardless of actual demand. As these capital expenditures grow, investors are increasingly seeking evidence that such spending will generate returns.Thu, 16 Jul 2026 16:27:37 GMTCompanies Fundinghttps://aisurfing.org/news/building-a-restaurant-telephony-ai-host-with-amazon-bedrock-agentcore-and-amazon-nova-2-so-02b5bd8a/https://aisurfing.org/news/building-a-restaurant-telephony-ai-host-with-amazon-bedrock-agentcore-and-amazon-nova-2-so-02b5bd8a/AWS published a guide to building a voice ordering system using Amazon Bedrock AgentCore and Amazon Nova 2 Sonic for real-time speech. The system connects to a restaurant backend via the Model Context Protocol (MCP) and uses a SIP gateway on ECS/Fargate to bridge phone calls. It pre-warms the agent session during ringing to avoid dead air.Thu, 16 Jul 2026 16:23:30 GMTProducts Agentshttps://aisurfing.org/news/google-renames-notebooklm-to-gemini-notebook-d7c6e13e/https://aisurfing.org/news/google-renames-notebooklm-to-gemini-notebook-d7c6e13e/Google is renaming its AI note-taking app NotebookLM to Gemini Notebook. The app will remain standalone but will integrate more deeply with Gemini and Google Search.Thu, 16 Jul 2026 16:19:27 GMTProducts Agentshttps://aisurfing.org/news/google-s-ai-mode-now-lets-you-link-and-interact-with-select-apps-8ec4902f/https://aisurfing.org/news/google-s-ai-mode-now-lets-you-link-and-interact-with-select-apps-8ec4902f/Google is expanding AI Mode beyond answering questions, enabling it to complete tasks across select apps that users regularly use. The update allows users to link and interact with certain applications directly through AI Mode.Thu, 16 Jul 2026 16:15:30 GMTProducts Agentshttps://aisurfing.org/news/interactive-proofs-for-general-distribution-properties-11eac62b/https://aisurfing.org/news/interactive-proofs-for-general-distribution-properties-11eac62b/Apple researchers have developed interactive proof systems that enable a verifier with limited samples to efficiently check claims about an unknown distribution made by an untrusted prover. These protocols apply to properties that can be decided by bounded-depth circuits and allow verification using fewer resources than independently running the analysis.Thu, 16 Jul 2026 15:39:28 GMTResearchhttps://aisurfing.org/news/soofi-consortium-releases-soofi-s-30b-a3b-open-hybrid-mamba-transformer-moe-model-for-germ-cbfcea46/https://aisurfing.org/news/soofi-consortium-releases-soofi-s-30b-a3b-open-hybrid-mamba-transformer-moe-model-for-germ-cbfcea46/The Soofi Consortium has released Soofi S 30B-A3B, an open-source hybrid Mamba-Transformer mixture-of-experts (MoE) foundation model. The model activates 3.2 billion of its 31.6 billion parameters and is designed for both German and English languages.Thu, 16 Jul 2026 15:35:33 GMTModelshttps://aisurfing.org/news/sakana-ai-s-orchestrator-adds-nvidia-nemotron-to-explore-collective-intelligence-versus-si-373e6bb5/https://aisurfing.org/news/sakana-ai-s-orchestrator-adds-nvidia-nemotron-to-explore-collective-intelligence-versus-si-373e6bb5/Sakana AI is integrating Nvidia's open-source Nemotron models into its Fugu orchestrator, which dynamically combines multiple language models for specific tasks. The company suggests that open models could become competitive with frontier systems when used in a coordinated way, though no specific benchmark results for this integration have been released yet.Thu, 16 Jul 2026 15:31:27 GMTModelshttps://aisurfing.org/news/moonshot-s-upcoming-kimi-3-aims-to-narrow-gap-with-anthropic-s-opus-4-8-3230f97b/https://aisurfing.org/news/moonshot-s-upcoming-kimi-3-aims-to-narrow-gap-with-anthropic-s-opus-4-8-3230f97b/Moonshot's Kimi K3 is expected to be the largest open AI model from China, with a parameter count between 2 trillion and 3 trillion. The model is anticipated to narrow the performance gap with Anthropic's Opus 4.8, according to recent reports.Thu, 16 Jul 2026 15:27:26 GMTModelshttps://aisurfing.org/news/openai-and-work-louder-unveil-codex-micro-a-hardware-controller-for-ai-agents-5fe2fb53/https://aisurfing.org/news/openai-and-work-louder-unveil-codex-micro-a-hardware-controller-for-ai-agents-5fe2fb53/OpenAI and keyboard manufacturer Work Louder have unveiled the Codex Micro, a compact hardware controller designed for interacting with AI agents. The device features a joystick, offering an alternative to typing commands for controlling AI workflows.Thu, 16 Jul 2026 15:23:27 GMTProducts Agentshttps://aisurfing.org/news/ami-labs-ceo-alexandre-lebrun-rejects-agi-and-superintelligence-labels-for-ai-895c7031/https://aisurfing.org/news/ami-labs-ceo-alexandre-lebrun-rejects-agi-and-superintelligence-labels-for-ai-895c7031/Alexandre LeBrun, CEO of AMI Labs, rejects the use of terms like 'AGI' and 'superintelligence' to describe his company's AI work. He believes these labels are overhyped and distract from meaningful progress in the field.Thu, 16 Jul 2026 15:19:31 GMTPeople Institutionshttps://aisurfing.org/news/brain-implant-helps-paralysed-man-to-feed-himself-and-drink-from-cup-97b82daf/https://aisurfing.org/news/brain-implant-helps-paralysed-man-to-feed-himself-and-drink-from-cup-97b82daf/A paralysed man regained the ability to move his arms and hands and feel touch after receiving a 'double neural bypass' brain implant. The technology, trialed since 2021, enabled him to feed himself and drink from a cup following surgery and months of training.Thu, 16 Jul 2026 15:15:26 GMTResearchhttps://aisurfing.org/news/google-ordered-to-open-android-and-search-to-rivals-in-europe-cc015f18/https://aisurfing.org/news/google-ordered-to-open-android-and-search-to-rivals-in-europe-cc015f18/European Union regulators have ordered Google to provide rival AI assistants and search engines with greater access to Android and Google Search, in line with the bloc's digital antitrust rules. The decisions, announced Thursday, are intended to prevent Google from using its Android user base to gain an unfair advantage in AI and search.Thu, 16 Jul 2026 13:27:26 GMTPolicy Safetyhttps://aisurfing.org/news/nebius-embarks-on-asset-light-data-center-model-059f7ffb/https://aisurfing.org/news/nebius-embarks-on-asset-light-data-center-model-059f7ffb/Nebius is adopting an asset-light data center model by partnering with infrastructure providers to expand its compute capacity without incurring the full costs of building and maintaining data centers. This approach enables Nebius to scale its operations more efficiently.Thu, 16 Jul 2026 13:23:30 GMTInfrastructurehttps://aisurfing.org/news/why-your-kubernetes-scheduler-can-t-handle-ai-workloads-61cf2767/https://aisurfing.org/news/why-your-kubernetes-scheduler-can-t-handle-ai-workloads-61cf2767/The default Kubernetes scheduler (kube-scheduler) lacks gang scheduling and multi-node fabric topology awareness, which can lead to partial-scheduling deadlocks and increased communication latency for distributed AI training jobs. These limitations can severely impact workflows for organizations running training across many GPUs, as jobs may be unable to start or run inefficiently due to poor scheduling decisions.Thu, 16 Jul 2026 13:19:26 GMTInfrastructurehttps://aisurfing.org/news/claude-can-now-use-your-1password-credentials-for-you-0337280e/https://aisurfing.org/news/claude-can-now-use-your-1password-credentials-for-you-0337280e/1Password has launched a browser integration for Claude, allowing the Anthropic chatbot to access stored credentials such as usernames and passwords. With user authorization, Claude can complete multi-step tasks like booking travel and managing online accounts without manual login input.Thu, 16 Jul 2026 13:15:28 GMTProducts Agentshttps://aisurfing.org/news/foundation-launched-to-standardize-ai-payments-4ce687d1/https://aisurfing.org/news/foundation-launched-to-standardize-ai-payments-4ce687d1/A new foundation has been launched to standardize AI payments and will oversee the x402 payments protocol. The initiative is supported by 40 members, including Visa, Mastercard, Google, and Microsoft.Thu, 16 Jul 2026 12:15:27 GMTInfrastructurehttps://aisurfing.org/news/google-deepmind-and-isomorphic-labs-share-joint-approach-to-bioresilience-and-ai-models-2dc6edf5/https://aisurfing.org/news/google-deepmind-and-isomorphic-labs-share-joint-approach-to-bioresilience-and-ai-models-2dc6edf5/Google DeepMind and Isomorphic Labs have published their joint approach to bioresilience, describing how they are using AI models to address biological risks. Their blog post outlines strategies for leveraging AI to enhance preparedness and response to biological threats.Thu, 16 Jul 2026 11:19:26 GMTPolicy Safetyhttps://aisurfing.org/news/the-problem-ai-content-moderation-cannot-solve-46901e4e/https://aisurfing.org/news/the-problem-ai-content-moderation-cannot-solve-46901e4e/Meta and other major tech companies are increasingly relying on AI for content moderation. However, as the backlash to Muse Image demonstrates, AI systems struggle to protect users because they do not account for issues of consent.Thu, 16 Jul 2026 11:15:25 GMTPolicy Safetyhttps://aisurfing.org/news/gemma-4-receives-stealth-update-fixing-tool-calling-and-truncation-bugs-29e915b9/https://aisurfing.org/news/gemma-4-receives-stealth-update-fixing-tool-calling-and-truncation-bugs-29e915b9/Google has quietly updated its open AI model Gemma 4, addressing bugs related to tool calling and truncated responses. The update also improves performance on Nvidia Hopper GPUs, while the model retains its original name.Thu, 16 Jul 2026 10:16:25 GMTModelshttps://aisurfing.org/news/xai-open-sources-grok-build-on-github-after-massive-data-breach-b959ec17/https://aisurfing.org/news/xai-open-sources-grok-build-on-github-after-massive-data-breach-b959ec17/xAI's command-line tool "Grok Build" was found to silently upload entire directories, including sensitive files like SSH keys and password databases, to Google Cloud servers. Following public backlash, Elon Musk pledged to delete all uploaded user data, and xAI subsequently open-sourced the full 844,530-line Rust codebase under the Apache 2.0 license.Thu, 16 Jul 2026 08:15:26 GMTPolicy Safetyhttps://aisurfing.org/news/heavy-tailed-flow-matching-via-random-clocks-f72708d2/https://aisurfing.org/news/heavy-tailed-flow-matching-via-random-clocks-f72708d2/Researchers introduce Heavy-Tailed Flow Matching via Random Clocks (HTFM), a framework that models heavy-tailed data by representing sources as mixtures of Gaussian distributions conditioned on random clock paths. The method demonstrates improved mode coverage, sample quality, and recovery of tail statistics on imbalanced datasets such as CIFAR10-LT and weather fields, while maintaining efficient sampling. HTFM also enables practical control over the heaviness of generated tails by adjusting the clock law or tail parameter.Thu, 16 Jul 2026 06:49:27 GMTResearchhttps://aisurfing.org/news/causal-analogical-researcher-cana-framework-enhances-llms-use-of-historical-analogies-for--94144760/https://aisurfing.org/news/causal-analogical-researcher-cana-framework-enhances-llms-use-of-historical-analogies-for--94144760/A new preprint introduces Analogical Deep Research (ADR), a task designed to evaluate large language model (LLM) agents on their ability to retrieve and integrate historical analogies for foresight analysis. The authors find that LLMs often rely on surface-level similarities rather than underlying causal mechanisms when identifying analogies. To address this, they propose the Causal Analogical Researcher (CANA) framework, which uses structural decomposition and feedback to improve analogy identification. CANA demonstrates up to a 10% improvement over previous methods on the ADR-bench benchmark.Thu, 16 Jul 2026 06:45:27 GMTResearchhttps://aisurfing.org/news/applied-computing-raises-20m-to-build-ai-foundation-model-for-oil-and-gas-plants-d4b4ace5/https://aisurfing.org/news/applied-computing-raises-20m-to-build-ai-foundation-model-for-oil-and-gas-plants-d4b4ace5/Applied Computing has raised a $20M Series A to develop a foundation AI model for the oil, gas, and petrochemical industry. The company aims to provide operators with an AI system designed for entire plant operations.Thu, 16 Jul 2026 06:41:25 GMTCompanies Fundinghttps://aisurfing.org/news/a-better-way-to-turn-2d-designs-into-3d-models-for-rapid-prototyping-83eac764/https://aisurfing.org/news/a-better-way-to-turn-2d-designs-into-3d-models-for-rapid-prototyping-83eac764/MIT researchers have developed an automated framework that helps AI models generate CAD programs from 2D designs more accurately and efficiently. This advancement could make it easier to convert 2D sketches into 3D models for rapid prototyping.Thu, 16 Jul 2026 06:37:25 GMTResearchhttps://aisurfing.org/news/algorithms-achieve-fixed-parameter-tractability-for-differentially-private-synthetic-data--ce5b63e7/https://aisurfing.org/news/algorithms-achieve-fixed-parameter-tractability-for-differentially-private-synthetic-data--ce5b63e7/A new preprint establishes that generating synthetic data under differential privacy is fixed-parameter tractable (FPT) when parameterized by the treewidth of the query family's incidence graph. The authors introduce two algorithms that achieve optimal error rates: one based on linear programming and the FPT of the LP dual's separation problem, and another using a subsampled private multiplicative weights method with FPT Gibbs sampling. Both approaches are unified by a dynamic programming framework over tree decompositions.Thu, 16 Jul 2026 06:33:28 GMTModelshttps://aisurfing.org/news/fast-rates-for-semi-supervised-learning-via-data-augmentation-graph-regularization-d7b8445b/https://aisurfing.org/news/fast-rates-for-semi-supervised-learning-via-data-augmentation-graph-regularization-d7b8445b/A new theoretical analysis demonstrates that data augmentation in semi-supervised learning can achieve a fast O(1/n_L) error rate with respect to the number of labeled samples, improving over the standard supervised O(1/√n_L) rate. The bound explicitly connects error to the quality of augmentations, measured by the graph-cut mass of augmentations crossing label boundaries. This work provides a mechanistic explanation for how augmentation quality influences the trade-off between accuracy and label count.Thu, 16 Jul 2026 06:31:34 GMTResearch