Search “Google”

Karpathy builds HackerTube with Claude Opus 5.5 in two hours

Andrej Karpathy spent the weekend using Claude Code together with Claude Opus 5.5 to build the video recommendation site HackerTube in about two hours. According to a report by A Yi AI Notes, he used Claude to analyze 43 million Hacker News comments from the past two decades and filtered out 6,454 long videos that are frequently recommended; the site requires no login, thumbnails are taken directly from the middle frame of each video, and browsing state is saved in the local browser session. He said he only manually applied for a Google Cloud YouTube API key, rented a VPS and bought a domain, after which he handed the IP to Claude, which completed deployment, security hardening and code optimization. A cnBeta report likewise mentions that the site publicly releases a ZIP package of the comment analysis data for others to adapt on their own. There is currently no further progress.

360's Tu Longfeng tops CyberGym with a 99.4% vulnerability reproduction success rate

360 Group's self-developed AI security attack-defense system Tu Longfeng topped the CyberGym benchmark with a 99.4% vulnerability reproduction success rate, competing on the same leaderboard against models and agents from vendors including OpenAI, Anthropic, and Google. The evaluation covered 1,507 tasks, and Tu Longfeng achieved success rates of 99.27%, 99.47%, and 99.40% across three independent rounds of evaluation, averaging 99.38%. 360 said Tu Longfeng has discovered more than 15,000 vulnerabilities in total, and over 500 enterprises have already connected to the system to conduct security testing.

Alphabet's third-quarter revenue and earnings per share both fell short of expectations

Alphabet reported third-quarter adjusted earnings per share of $2.67 and revenue of $127.58 billion, both below analysts' expectations. According to the Google Blog, revenue for the quarter grew 25% year over year, with annualized revenue reaching $500 billion, and Google Cloud revenue grew 103%; CEO Sundar Pichai said the first-party model API processes nearly 32 billion tokens per minute, 40% more than the previous quarter, and the Gemini app has reached 1 billion users. Solidot reported that the stock price fell nearly 10% after the earnings release, and the company disclosed that revenue generated by AI search is lower than that of traditional search, with overall search business revenue growing 12% year over year to $63.25 billion; Chief Business Officer Philipp Schindler said the slowdown in search business growth stems from the lower monetization capability of AI Mode. Spending for the quarter was $51.4 billion, and net profit was $26.17 billion, lower than $35 billion in the same period last year.

Rohan Paul said that Sunra resells multimodal model calls at the manufacturer's original price

Rohan Paul stated on October 29, 2026 that Sunra has released the Sunra API, through which users can call all major image, video, audio, and 3D models with a single key, billed at each model vendor's own API prices on a pay-as-you-go basis, with no subscription fee and no markup. Rohan Paul also said that Seedance 2.5 is priced on Sunra at half of what two other gateways charge, and that this unified entry point also makes it easier to run A/B tests on models from Google, OpenAI, ByteDance, and Kling. At present, this claim appears only in Rohan Paul's account, and there is no independent evidence from Sunra officially or from other sources.

MiniMax Code team open-sources Agent workflow and Skills

On October 28, 2026, the MiniMax Code team open-sourced MCode Stack, packaging the Agent workflows and Skills the team uses daily into installable SKILL.md files and manifests, which can be integrated as plugins into MiniMax Code and any Agents that support Skills. The repository packages stacks by member, among which stacks/ronny/ is the most complete, containing 15 skills covering sub-Agent decomposition, context and tool budget auditing, long-session handoff, independent review, and hotfix judgment. An accompanying notes/ column, MCode Engineering Notes, explains the rationale; six engineering-* entries are adapted from Google Engineering Practices, with the adapted portions retaining CC-BY 3.0 and the rest under MIT. The next day, reports stated that the project is open-sourced under the MIT license, includes 17 SKILL.md skills and 5 manifests, does not contain MiniMax Code source code, and can be installed by copying the skill directories, via a local plugin, or through a plugin marketplace, and can also be installed in Claude Code; the original text did not publish evaluation results.

Pichai says AI is helping improve training and release speed

On October 29, 2026, Google CEO Sundar Pichai said on an earnings call that AI is helping improve training and release speed, and he expects this to continue accelerating over the next twelve months due to RSI. Related reports suggest this is not equivalent to an official announcement of recursive self-improvement, but AI feeding back into the R&D pipeline has already been written into investor communication records; the real change is that engineering iteration cycles are being compressed, and the next stage will be a contest over who can hand over the R&D pipeline to AI faster. At present, this claim comes from a single report, with no further independent sources or official materials for cross-verification.

Xreal Aura has officially launched in four countries, starting at $1,099

On October 28, 2026, Google and XREAL released the XREAL AURA, the first wired XR glasses powered by Android XR, starting at $1,099. According to the Google Blog, the product is available for order starting today in the United States, Canada, and South Korea, and goes on sale in Japan on November 2. The glasses weigh less than 95 grams, use transparent lenses and a 70-degree field of view to project apps into the user's field of vision; heavy computing is handled by a portable computing puck connected via cable, which features the debut of the Qualcomm Snapdragon Reality Elite platform. Users can open multiple apps at once and pair them with a Bluetooth keyboard and mouse, or connect a laptop screen as a window via USB-C wired mirroring, and Gemini can provide assistance with content in the field of view and on the screen. Testing Catalog reported the same day that the product is aimed at users aged 18 and over, with a launch price starting at $1,099 within the year.

Gemini Notebook offers over 700 textbooks for free

On October 29, 2026, Google announced that Gemini Notebook (formerly NotebookLM) has added hundreds of textbooks from U.S. and Indian publishers such as OpenStax, Open Textbook Library, and S Chand Publishing, which students can add for free. Google's Josh Woodward said on X that more than 700 textbooks are available for free. Users can select Add sources in a notebook and then choose Textbooks to view and add them, filtering by subject, level, and country. After adding textbooks, every answer is directly grounded in the textbook content, with inline citations and links to the original text, and the textbook content can also be turned into study guides, practice questions, and flashcards.

Anthropic upgrades Claude mobile iPad and iPhone Duo interface

On October 30, 2026, Anthropic upgraded the Claude mobile app; on iPad and iPhone Duo, generated content now opens alongside the chat. In the same batch of updates, the iOS and Android versions support simultaneous login to multiple accounts with switching without logging out; Chromebook, Googlebook, and foldable devices gain an adjustable sidebar, touchpad and right-click support, and keyboard shortcuts; voice mode on iOS and Android responds faster and recognizes speech more accurately in noisy environments, and iPhone voice mode is integrated with Apple CarPlay. Phoenix Technology's report also mentioned that Claude added 24 new languages, bringing cumulative support to 33 languages.

Gemini offers an immersive exploration feature

Gemini app adds an immersive exploration library; users can click "Try immersive explorations" in the student hub to enter and explore topics from microscopic systems to deep space layer by layer in an interactive, visual way. According to the Google Blog, each scene has clickable labels for deeper viewing and will soon support audio narration; the feature will later also be integrated directly into conversations with Gemini. The feature supports English, Indonesian, Hindi, Brazilian Portuguese, French, and Simplified Chinese; users of other languages can currently only use the English version. Other reports mention that users can explore topics such as Silk Road trade routes, the interior of a jet engine, and the physics of lightning, with each scene unfolding in interactive layers accompanied by guided narration.

Nathan Lambert criticized the Gemini model for publishing scores despite not being publicly available

On October 29, 2026, Nathan Lambert publicly criticized Google DeepMind's Gemini model: he stated that every time he wanted to try the model, it was not publicly accessible, and he believes that if the outside world cannot use it, there is no need to publish benchmark scores. Currently, this criticism appears only in an article he himself published, and there has been no response from Google DeepMind or other subsequent developments.

Gemini app launches in-chat file generation and editing

On October 30, 2026, the Gemini app launched the ability to generate, preview, and edit files within chats. According to the Google Blog, users can type edits directly in full-screen mode within a conversation, select text to ask questions about it, or have Gemini rewrite the file; PDF preview and editing are supported on both web and mobile, while other formats such as Google Docs, Google Sheets, Slides, Markdown, TXT, and code are supported only on web. The feature is available to all users in all languages covered by the Gemini app.

Gemini launches avatar creation and sharing feature

Google Gemini has launched a personal avatar feature, allowing users to create and share avatars and appear together with friends in images and videos. After sharing, users can tag each other in prompts and maintain consistency of characters and voices; users can also add or remove others' access at any time. Currently, this feature is based on information disclosed by Gemini on October 30, 2026, and there are no further developments yet.

Gemini adds the ability to connect Strava accounts

Gemini adds the ability to connect Strava accounts, which can turn users' years of raw training logs into personalized coaching plans. According to information posted by Josh Woodward on October 29, 2026, this feature allows Gemini to read Strava training data and generate personalized training plans. The next day, 9to5Mac reported that Strava announced this week it is expanding its AI integration to Google Gemini, having previously connected to Claude via MCP; after connecting, users can query their training data in natural language, for example comparing this year's fastest 10K with historical bests, analyzing mile-by-mile race performance, judging the impact of weather on long runs based on temperature, humidity and wind, and viewing their longest streak records across various sports. The feature is currently public along with the above reports.

Gemini 4 Carbon Tops the Artificial Analysis Intelligence Index

On October 28, 2026, Google DeepMind released Gemini 4 Carbon, calling it its strongest model, ranking first with 59 points on the Artificial Analysis Intelligence Index v5.0, and scoring 88.2% on SEC-bench Pro, 54.8% on ExploitGym, and 71.0% on CWE-bench. The model is aimed at long-horizon engineering tasks and was first opened to some cybersecurity defenders before being expanded through the Fairwind Program. On October 29, Arena.ai said Gemini 4 Carbon ranked 2nd on Agent Arena with a net improvement of 16.2% and pushed the Pareto frontier forward; the same day, Yuchen Jin cited Artificial Analysis data to raise reservations, noting that the evaluation used the Antigravity CLI, and he planned to try it instead in Codex and Omnigent. On October 30, the Artificial Analysis Intelligence Index showed Carbon in first place with 59 points, 1 point higher than Claude Opus 5.5. On October 31, Perplexity CEO Aravind Srinivas shared an early evaluation, saying it set a new SoTA on the WANDR benchmark. There is also news that

View story:Google releases the Gemini 4 Carbon model

Alex Krentsel presents the self-improving AI agent architecture Exo at Arena HQ

Alex Krentsel, a senior researcher at Google Research and a PhD student at UC Berkeley, shared the self-improving AI agent architecture Exo at Arena. The architecture draws on the exokernel approach from operating systems, splitting the agent into a protected mechanism layer (state, history, keys, and a "guardian") and a fully evolvable policy layer. On benchmarks and open-world RuneScape tasks, Exo is 31-62% cheaper than Codex and Pi. Currently, this presentation has only one report from Arena, with no follow-up developments yet.

Intel announces the acquisition of Classiq for $2.3 billion

Intel announced the acquisition of the Israeli quantum software company Classiq for $2.3 billion, with the deal expected to close in the second half of next year. According to The Next Platform, Classiq's platform uses AI natural language models and a Python-like language to help developers build quantum algorithms and applications, and it already has more than 40 enterprise customers, including Comcast, AT&T, BMW, and Oak Ridge National Laboratory. Intel CTO Pushkar Ranade said the move will turn its quantum research team, which has existed for 10 years, into a real quantum computing business, making Intel a full-stack quantum company able to compete with IBM and Google. No further progress has been disclosed at this time.

Akamai Launches Beta Versions of Akamai Inference and Agent Sandbox

On October 29, Akamai released beta versions of two products, Akamai Inference and Akamai Agent Sandbox, allowing users to run AI models on Akamai Cloud and securely execute code generated by AI agents. According to The Elec, Akamai Inference supports 12 open-weight models, including OpenAI's gpt-oss-120b, Google's Gemma 4, and Nvidia's Nemotron 3 Ultra, and provides an OpenAI-compatible API; Agent Sandbox is based on WebAssembly isolation technology and restricts code access to computing resources, network connections, and system functions. Analytics India Magazine separately reported that Agent Sandbox is released as an open-source component of the Spin framework, and that Akamai is also expanding its NVIDIA Blackwell Ultra GPU cluster, with each eight-GPU node equipped with 2.3 TB of HBM3e memory. Both services are currently in the beta stage.

Gemini supports editing AI-generated files within the chat window

On October 29, 2026, Google added the ability to directly create, preview, and edit files to the Gemini app and web version, supporting documents (text, Markdown, LaTeX), PDFs, presentations, and spreadsheets. Users can select content and have Gemini modify text or visual elements, and can continue asking follow-up questions to make adjustments; generated files can be downloaded in their original format, opened in apps such as Google Slides, or uploaded to Google Drive via the export button in the upper right corner. According to Thurrott, editing in the chat window for documents, PDFs, presentations, and spreadsheets is already available on the web and mobile, while editing for Google Docs, Google Sheets, Markdown, TXT, slides, and code is available only in the Gemini web app. PDF preview and editing are supported on both web and mobile, while other formats are supported only on the web; the feature is available to all languages and users who can use the Gemini app.

Oracle announces the introduction of Cerebras AI chips into its cloud services

Oracle announced that it will introduce Cerebras Systems' AI chips into its cloud infrastructure, giving customers more computing power options beyond Nvidia GPUs. According to Data Center Knowledge, Oracle also plans to deploy Google TPUs and inference chips from OpenAI, Cerebras, SambaNova and Positron AI on OCI, with the TPU and Cerebras wafer-scale engine scheduled to come online in 2027, and Positron Atlas already available. CNBC Technology reported the same day that Alphabet announced Oracle will offer Google's TPUs, marking the first time customers can use TPUs outside of Google's infrastructure. No further developments have been disclosed at this time.