English text is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

Hygon Information Discloses 2026 Third Quarter Report

Hygon Information disclosed its 2026 third quarter report in October 2026. According to Jiwei Network, Hygon Information achieved operating revenue of 14.955 billion yuan in the first three quarters of 2026, a year-on-year increase of 57.58%; net profit attributable to the parent company was 2.847 billion yuan, a year-on-year increase of 45.12%. Among them, single-quarter revenue in the third quarter was 5.856 billion yuan, a year-on-year increase of 45.46%. R&D investment in the first three quarters was 4.472 billion yuan, accounting for 29.90% of revenue, a year-on-year increase of 52.39%. Currently, the relevant data in this report is reported by Jiwei Network as a single source, with no other follow-up developments for the time being.

English translation is being prepared.

Apple launches HomeView smart home display

On October 13, 2026, Apple launched the Apple HomeView smart home display, featuring a 7-inch Liquid Retina display and Siri AI powered by Apple Intelligence, capable of recognizing up to six users and providing personalized content, with support for FaceTime, music playback, and smart home control. Both the HomeView and the wall-mounted version are priced at $299, with pre-orders opening that day in Australia, Canada, India, New Zealand, the UK, and the US, and shipping on November 13; because Siri AI is not yet available on iOS in the EU, the product is not sold in the EU. On October 14, Apple updated its developer agreement so that iOS app widgets can be offered on CarPlay and HomeView, allowing this device, which has no App Store, to display third-party widgets. Reports on October 17 said that for users who did not pre-order, some color options are expected to wait until early December, with the silver desktop version in New York at November 23 to December 1, and in other regions at November 30 to December 7, while the wall-mounted version currently has no delay.

OpenAI blog proposes two complementary monitorability metrics

On October 20, 2026, OpenAI published a post on its Alignment research blog proposing two complementary monitorability metrics in addition to g-mean2. According to the post, the pair-matched metric compares monitorability only on instances that are jointly supported by multiple models; the post-control risk metric simulates the probability that the system still produces harmful outputs after the monitor rejects an output. The post notes that g-mean2 was previously used to measure a monitor's ability to distinguish positive from negative samples, but because different models have different subsets of valid positive-sample tasks, the scores may reflect different mixtures of tasks. The post also states that if the monitor is used as a deployment control, the model can retry after being rejected, and the risk depends on the frequency of bad behavior, which outputs the monitor accepts, and the number of attempts allowed. This content currently comes from OpenAI's official blog and has not yet been cross-verified by other independent sources.

xAI adds two incident-handling AI teammates to Grok Bot

xAI has added two AI teammates, IncidentWatch and IncidentSweep, to Grok Bot, responsible respectively for monitoring incident signals in customer service tickets and for closing affected tickets in bulk. According to information published by xAI News, IncidentWatch checks tickets from the past 15 minutes every 5 minutes and notifies the on-duty support engineer when it detects a surge of similar issues. In one incident on September 18, IncidentSweep cleared about 204 tickets in roughly 10 minutes with the review and approval of a support engineer. This information is currently published by xAI alone, and there are no follow-up developments from other independent sources yet.

Artificial Analysis says the cost of reaching intelligence index thresholds is falling rapidly

On October 20, 2026, Artificial Analysis put forward the judgment that the cost of reaching each score threshold on the Intelligence Index is falling rapidly. Based on v4.3.2 measured results and current pricing statistics, it stated that the per-task cost of reaching 30 points is 86 times lower than when that score was first reached: GPT-6 Luna (high) reaches 30 points at $0.03 per task, whereas Claude Sonnet 4.6 cost $2.49 per task when it first reached 30 points in February 2026. The same statistics also stated that the cost of reaching 20 points is 123 times lower than about 13 months ago, and the cost of reaching 40 points is 33 times lower than about 5 months ago; Claude Haiku 5.5 (xhigh) reaches 41 points at $0.12 per task, making it the cheapest model to reach 40 points. The highest measured score rose from 21 to 58 within a year, at a cost of about 11 times, with Claude Opus 5.5 at $5.98 per task. At present, this judgment is published by Artificial Analysis as a single source, and no other independent source has cross-verified it.

n8n Unveils Brand Refresh at In The Loop Conference

On October 14, 2026, n8n unveiled a new logo, new typeface, and adjusted color system at the In The Loop conference, with pink remaining central. According to the n8n Blog, the new logo shifts from emphasizing "connecting nodes" to expressing the activity that happens between nodes, echoing n8n's evolution from workflow automation toward responsive, action-capable agent systems. The same source states that the Community Edition, model selection, and community ecosystem remain unchanged. This news was published by the n8n Blog on October 20, and no other independent sources have followed up yet.

Schaeffler opens humanoid robot joint module factory in Taicang

Schaeffler has opened a production factory for humanoid robot joint modules in Taicang. According to a report by heise online on October 20, 2026, the factory is called Lighthouse Factory, covers an area of about 7,500 square meters, and uses 6 humanoid robots at 5 workstations to produce linear and rotary actuators for use in other humanoid robots. The factory has an assembly line, a training and data collection center, a digital twin platform, and a data center, and uses the digital twin to virtually map and pre-test the production process. The report also mentioned that Schaeffler previously announced plans to deploy hundreds of humanoid robots in its own factories over the next five years, and reached an agreement with Humanoid to supply four-digit numbers of robots by 2032. At present, this event has only the above-mentioned single-source report, and there are no subsequent factual developments.

Everpure launches Data Intelligence data management software

On October 20, 2026, Everpure launched its Data Intelligence data management software. According to a report by heise online the same day, the software is used to identify and classify structured and unstructured data across storage locations and to supplement it with metadata and semantic context so that AI agents can use it. The report said it supports SMB and NFS file shares and S3 object storage, and offers nearly 300 connectors that can connect to other vendors' storage systems, SaaS applications such as Salesforce and ServiceNow, and mainframes, without the need to centrally migrate data. The report also mentioned that the newly added File Intelligence can analyze the access permissions and file age of file shares, and that the newly integrated MCP lets AI agents query the data catalog, but they do not automatically gain permission to read the corresponding data. There are currently no subsequent developments.

Scaled Cognition launches voice model Voice

On October 20, 2026, Scaled Cognition Inc. launched its text-to-speech model Voice. According to a SiliconANGLE report the same day, the model supports more than 25 languages and voice cloning, can be deployed via API calls or self-hosted, and focuses on accurate reading of details such as account numbers and addresses. The report said hosted API pricing is $9 per million characters, while self-hosting is $1 per million characters (excluding infrastructure); the company says that at a scale of 5 million characters, including infrastructure, the cost is about $13 per million characters, and it claims a speech generation latency of 70 milliseconds. Its Smart Pronunciation capability first determines the meaning of abbreviations and numbers before generating audio—for example, reading "Pay 1/2 by 1/2" as "pay half by January second"—and it can run without relying on APT or other large language models. There are currently no subsequent developments.

AYi said after a real-world comparison test that domestic models have risen

On October 20, 2026, AYi AI Notes released a real-world comparison test, testing MiniMax M3.1 Flash and Claude Opus 5.5 side by side. It said that with one prompt and about 1.5 hours, it built a 3D interactive solar system webpage, with results comparable to Opus 5.5 or even better overall. AYi thus concluded that domestic models this time give people hope, and said that its previous criticism of domestic models deserves an apology. Currently this claim comes from its personal testing and evaluation, with no cross-verification from other sources.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

Nebius Expands AI Builder Program to Europe

Nebius expands its AI Builder Program to Europe. According to a Testing Catalog report on October 19, 2026, the program partners with NVIDIA, Accel and other European venture capital firms to provide a combined $10 million in compute grants, funding and community program support. The program is free to join; developers can receive credits for running open-source models on Nebius, and selected projects can also receive compute grants at Token Factory. The European edition additionally supports larger open-source projects that need more compute or funding, and hosts hackathons and demo days with communities such as AI Tinkerers, {Tech: Europe} and ESRA. There are currently no subsequent developments.