All news

CcnBeta01:59

AI score64

Google says search and AI can no longer be separated; schools that ban AI will lose Google Search

Google announced plans that, starting in 2027, Gemini, NotebookLM, Gemini in Classroom, and Google Search-related AI features in school-managed K-12 student accounts will be disabled by default and will require school administrators to actively enable them. Chris Phillips, Vice President of Google's education business, said that search and AI are now deeply integrated, and that if schools completely ban AI, students will no longer be able to use Google Search and will only see a disabled notice set by administrators. Google also announced the promotion of Guided Learning, a new AI learning tool in education, which replaces direct answers with guided questioning.

Same story · 15 reports

#Industry news#Google#English text is being prepared.

HHugging Face01:57First-party

AI score46

CWE-bench v2: Comprehensive improvements in coverage, quality, and transparency

CWE-bench v2 publishes evaluation results: Among 11 models, Gemini 4 Argon leads with 67.8% pass@1 and 81.0% pass@4, followed closely by Claude Opus 5.5 with 67.3% pass@1. 95 tasks were solved by at least one model, but the top model still fails 19% of attempts. On cost, DeepSeek V4.1 Flash, with 57.0% pass@1 and $0.04 per rollout, forms the Pareto frontier together with Claude Opus 5.5 and Gemini 4 Argon; the dataset, scorer, and harness are all publicly available and can be run directly from the Hub via Harbor.

#Evaluations / benchmarks#Safety / alignment#Open-source ecosystem

GGoogle DeepMind01:57First-partyProduct release

AI score64

Google DeepMind launches Cheminformatics Skills for Antigravity

Google DeepMind has released a set of Cheminformatics Skills for Google Antigravity. Built on RDKit and connected to ChEMBL and PubChem, the language model handles task orchestration and scientific reasoning, while deterministic chemistry tools perform molecular computations. Using lung cancer drugs as an example, the agent searched thousands of compounds in about one hour, completed matched molecular pair analysis, trained a potency prediction model, and also proposed novel rigidified macrocyclic molecular designs with synthetic routes. The team also noted that sparse data, inter-laboratory assay noise, and narrow dynamic ranges limit prediction accuracy; computational designs are only verifiable hypotheses and still require experimental synthesis for validation. The skills are now live on Antigravity and can be used by installing the Science plugin, and free access to the highest usage tier for six months is being offered to 10,000 scientists.

Same story · 2 reports

#Product updates#Agents#English text is being prepared.

HHacker News Popular01:56

AI score40

JIVX: Turn YouTube Japanese Videos into Sentence-by-Sentence Practice

JIVX can split the dialogue of Japanese YouTube videos into sentences, allowing users to practice typing or shadowing and receive a score for each sentence; users can also directly use videos that others have already added. The practice page labels the number of sentences by N5 to N1; for example, N5 has 44 sentences and N4 has 4 sentences, covering topics such as daily life, culture, and travel.

#English text is being prepared.

HHacker News Trending01:56

AI score38

QUILI.AI: Residents of the Chilean town of Quilicura become "human AI" for a day to protest data center water consumption

More than 40 residents of the Chilean town of Quilicura became "human AI" from 8:00 to 20:00 on January 31, manually answering more than 25,000 prompts to protest the heavy water consumption of local AI data centers. The campaign covered 68 countries and reached more than 1 billion people worldwide. Quilicura is located in the Maipo River basin and has been home to a large concentration of AI data centers since the early 2010s, which use local water to cool servers, while Chile is facing a severe drought expected to last until 2040.

#English text is being prepared.#English text is being prepared.

HHacker News Trending01:56

AI score29

ByteDance USDB: Scaling Storage for Next-Generation Recommendation Systems

ByteDance proposes the unified sequence storage system USDB, which uses CascadeLSM to materialize user history in time-based tiers, shares persistent files across differentiated replicas, reconstructs inference inputs from partial dumps, and reuses maintenance results for writes, inference, and training. Production migration records show an 85.8% reduction in average write latency and a 42.1% reduction in normalized deployment cost; compared with an optimized RocksDB sequence baseline, long-history reads are faster, and aggregation pushdown achieves 8.5x the throughput of ordinary reads under a constrained receive budget.

#Papers/Research#English text is being prepared.#English text is being prepared.

Hheise online01:53

AI score25

Darknet Diaries German edition: Maxie — Between the Network and the Seabed (Part 2)

heise online has published Part 2 of the German edition of Darknet Diaries, "Maxie — Between the Network and the Seabed," recounting Maxie Reynolds' transition from the field of underwater robotics to penetration testing. During an authorized red team test at a high-security logistics company, with a colleague temporarily absent, she conducted a solo nighttime reconnaissance of the perimeter and found that the old wire fence could be lifted up, but that none of the doors or windows could be opened.

#English text is being prepared.

Hheise online01:53

AI score26

German Digital Summit: Retraining Everyone for Skilled Trades Is Not the Solution to AI-Driven Job Losses

At the German federal government's digital summit, labor market researcher Enzo Weber opposed treating "everyone going into skilled trades" as the solution to AI-driven job losses, advocating instead for vocational continuing education. In an Adaptavist survey, nearly three-quarters of knowledge workers in Germany said they wanted to return to their pre-ChatGPT working state. Weber also noted that AI overall will not reduce labor demand, that Germany still needs immigration, and mentioned that the number of radiologists has increased by 17% over the past decade.

#Industry news#English text is being prepared.

Hheise online01:53

AI score49

India proposes DEPA data platform concept: enabling AI models to legally access hospital and bank data

India's Ministry of Electronics and IT has proposed the "Data Empowerment and Protection Architecture" (DEPA): hospitals, banks, government agencies and other "Training Data Providers" hand data over to a central "Data Exchange Plattform" (DEP), where it is anonymized and encrypted before being made available to AI companies as "Training Data Consumers" to train models under the platform's rules. The concept is primarily aimed at India's domestic AI models, but does not exclude foreign users, and is said to benefit public healthcare, education and bank lending. India has previously funded projects such as the Unified Payments Interface and Bharat Taxi to safeguard digital sovereignty.

#English text is being prepared.#English text is being prepared.#Industry news

Hheise online01:53

AI score58

Palantir expects revenue of about $2.65 billion this quarter, up about 90% year over year

When Palantir reported earnings on Monday, it projected revenue of about $265 million this quarter, up about 90% year over year; in the previous quarter, revenue rose 96% year over year to $2.3 billion, and quarterly profit rose 49% year over year to $1.13 billion. Growth came mainly from the U.S. market, where U.S. government business grew 96% to $953 million and U.S. commercial business grew 148% to $984 million. The United Kingdom signed a $1 billion military products contract with Palantir in September 2025, and German Interior Minister Dobrindt is also evaluating the full use of its software in German security agencies.

Same story · 2 reports

#Industry news#English text is being prepared.

Hheise online01:53

AI score31

Commentary: Merz's empty sovereignty promises

German Chancellor Friedrich Merz defended the German intelligence agency's use of Google Cloud at the Berlin Digital Summit, saying the data "is secure for us" and stressing that Google is also investing in Europe. The commentary notes that the real risk is not forced US data access, but the "soft kill-switch" of software stack updates being in the hands of US vendors: once updates are cut off, only a few months remain to migrate. The author argues that without anchor customers like the German intelligence agency, Europe's independent cloud infrastructure would be difficult to both develop and operate, and that Merz's remarks undermine digital sovereignty efforts.

#English text is being prepared.#Industry news

Hheise online01:53

AI score16

BitWise: A Fun Electronic Circuit Puzzle Game on Steam

BitWise, an electronic circuit puzzle game developed by Faint Current, launched on Steam on October 13, priced at 12.25 EUR, and supports Windows, macOS, and Linux (including SteamOS). The game has a total of 238 levels (70 of which are optional), divided into seven chapters that progress from lighting a bulb with a battery to logic gates, adders, ALUs, and even a small computer capable of executing instructions; the analog circuits are calculated using the same nodal analysis method as SPICE. The author believes its basic concepts are suitable for beginners, but the transistors lack details such as base resistors, which makes them differ from real circuits, and the judgment in some levels depends on an accurate understanding of the wording of the problem.

#English text is being prepared.

UUnite.AI01:53Model releases

AI score61

Cogent releases VR-1.1, achieving a 57.9% fix pass rate on LiveCVEBench

Cogent Security released the vulnerability-fixing model VR-1.1 on November 3, 2026, achieving a 57.9% pass fix rate across 190 tasks in LiveCVEBench, compared with 45.3% each for Claude Code and Codex. The benchmark covers 14 programming languages and 153 repositories, and requires patches to pass both vulnerability checks and functionality checks. Cogent says the results come from security-fix training and a process of first reproducing the vulnerability, then applying a patch and detecting bypasses, while also noting that the baseline models' training cutoff dates are later than the publication dates of the relevant CVEs, and that production deployment still requires review and testing in the target environment.

Same story · 2 reports

#Model releases#Safety / alignment#Evaluations / benchmarks

UUnite.AI01:53

AI score58

CrowdStrike Plans to Acquire Fiddler and Dreadnode, Expanding Its AI Security Product Line

On November 3, 2026, CrowdStrike announced plans to acquire Fiddler and Dreadnode, with both deals unveiled at Fal.Con EU 2026. Fiddler's technology will become Falcon Observe, used to evaluate and explain the behavior of AI agents; Dreadnode, meanwhile, will add automated AI security testing to the Falcon platform, covering more than 50 attack techniques including jailbreaks, prompt injection, and multi-turn attacks, and aligning with the OWASP Top 10 for LLMs, MITRE ATLAS, and the NIST AI Risk Management Framework. CrowdStrike plans to make Falcon Observe and Dreadnode generally available in the fourth quarter of fiscal year 2027, and Dreadnode's test results will also feed into Falcon Guardian as policy recommendations.

Same story · 2 reports

#Industry news#Safety / alignment#Agents

UUnite.AI01:53Product release

AI score58

Snowflake Launches Always-On Background Agents for CoCo and CoWork

On November 3, 2026, Snowflake announced always-on background bots for CoCo, the coding agent for data teams, and CoWork, the personal work agent for knowledge workers. After users set high-level goals, the bots execute in the background within inherited permission scopes, prompting only at key decisions or final approval. Both bots are entering private preview soon; CoCo Automations is now generally available, supporting scheduled recurring workflows while retaining audit records and task history, and the Async Agent API is also generally available. Snowflake said that in the quarter ending July 31, 2026, over the four-week average more than 5,800 accounts used CoWork and more than 9,100 accounts used CoCo, with Cisco, Docusign, and United Rentals among the users.

#Product updates#Agents#English text is being prepared.

UUnite.AI01:53

AI score32

LexisNexis Launches Accurint One: Integrating Investigative Data and Governed AI for Law Enforcement Agencies

LexisNexis Risk Solutions has released Accurint One, providing a unified investigative environment for U.S. criminal justice agencies that integrates company, partner, and agency-owned data, advanced analytics, and governed AI. The platform features three built-in information pools and supports record summaries, relationship network analysis and timelines, maps, and link analysis, with AI playing only a supporting role while interpretation and action remain the responsibility of authorized users. Its responsible AI principles are governed by the five TRUST criteria, covering transparency, explainability, and traceability.

#Product updates#Safety / alignment#English text is being prepared.

UUnite.AI01:53

AI score27

Vertafore Launches Velocity Change Request Agent on Surefyre

On November 3, 2026, Vertafore announced that the Velocity Change Request Agent is now available in Surefyre for MGAs, with internal testing showing a 70% reduction in time spent reviewing and documenting changes and a 98% reduction in errors. The agent compares renewal and endorsement requests against existing policy information and presents the differences in plain language, while underwriters remain responsible for reviewing, editing, and saving the summary. It joins five Velocity AI agents—Insurance Expert, Portal Launcher, Submission Processing, Document Intelligence, and Configuration—and the Configuration Agent remains in the Early Adopter stage.

#Product updates#Agents#English text is being prepared.

UUnite.AI01:53

AI score37

ARIA Selects Lumai Optical Server for Rack-Scale AI Inference Testing

The UK ARIA-funded Scaling Inference Lab will deploy the Lumai Iris optical inference server as its first optical computing platform, testing the architecture at rack scale in a commercial data center. Lumai says Iris Nova supports 8B and 70B parameter Llama models, uses a hybrid optical-digital architecture focused on the prefill stage, and claims it can reduce data center power consumption by up to 90% and improve energy efficiency per inference by at least 16x, but these figures have not yet been independently verified by ARIA.

#Industry news#English text is being prepared.#Reasoning

UUnite.AI01:53

AI score83

Binance Discloses $634 Million Theft Incident: Fake Interviews and Internal AI Tool Compromised

In preliminary investigation findings published on November 3, Binance said that fake recruitment interviews, malicious developer software, and a compromised internal AI tool together led to the theft of approximately $634 million in digital assets on October 29. Binance said it will bear the losses itself, has set aside coverage funds exceeding 100%, had restored 98% of affected token assets as of October 31, and has frozen or intercepted approximately $285 million. The attackers also used Ethereum's EIP-7702 authorization mechanism to establish malicious delegations, making it impossible to immediately revoke on-chain permissions by suspending withdrawals; Binance has rebuilt the affected wallet system and internal AI tool in a new environment and will bring in independent third-party security experts for review.

Same story · 3 reports

#Safety / alignment#Industry news#Agents

UUnite.AI01:53Research

AI score73

Google opens 85 engineering benchmark PLE tasks and agent competition

Google AI & Infrastructure announced the release of 29 public tasks, scoring tools, and evaluation containers for Physical-Engineering's Last Exam (PLE), and launched an agent competition and leaderboard. PLE was developed by Google in collaboration with Mercor and comprises 85 mechanical, electrical, civil, and structural engineering tasks, with each task containing an average of 61 files. On the 56-task hidden test set, all models scored below 50% in their native harnesses, with the highest being Claude Opus 5.5 in Claude Code at 48.3%; on all 85 tasks, this combination scored 50.6%, consuming an average of 27 million tokens and 63 minutes per task, at a list-price cost of $17.18.

#Evaluations / benchmarks#Agents#Google