Futures
Access hundreds of perpetual contracts
TradFi
Gold
One platform for global traditional assets
Options
Hot
Trade European-style vanilla options
Unified Account
Maximize your capital efficiency
Demo Trading
Introduction to Futures Trading
Learn the basics of futures trading
Futures Events
Join events to earn rewards
Demo Trading
Use virtual funds to practice risk-free trading
Launch
CandyDrop
Collect candies to earn airdrops
Launchpool
Quick staking, earn potential new tokens
HODLer Airdrop
Hold GT and get massive airdrops for free
Pre-IPOs
Unlock full access to global stock IPOs
Alpha Points
Trade on-chain assets and earn airdrops
Futures Points
Earn futures points and claim airdrop rewards
Promotions
AI
Gate AI
Your all-in-one conversational AI partner
Gate AI Bot
Use Gate AI directly in your social App
GateClaw
Gate Blue Lobster, ready to go
Gate for AI Agent
AI infrastructure, Gate MCP, Skills, and CLI
Gate Skills Hub
10K+ Skills
From office tasks to trading, the all-in-one skill hub makes AI even more useful.
GateRouter
Smartly choose from 40+ AI models, with 0% extra fees
Babbitt | Daily must-read for the Metaverse: Kuaishou’s self-developed Wenshengtu large model “Ketu” has started internal testing; the Ministry of Industry and Information Technology will give priority to the development of industry application standards such as “Metaverse + Industrial Manufacturing”
Abstract: Following the launch of the large language model “KwaiYii” last month, Kuaishou launched the large Vincent graph model “Kolors” in less than a month, continuing its large model business. After “Wen Shengwen”, the territory of “Wen Shengtu” was added. According to a report by Future Science and Technology on September 18, Kuaishou’s latest progress in the field of AIGC, the self-developed Vincent graph model “Kolors”, has been fully launched within the company for internal testing.
Hot info:
Ministry of Industry and Information Technology: Prioritize the development of industry application standards such as “Yuanverse + Industrial Manufacturing”
Babbitt News, on September 18, the Department of Science and Technology of the Ministry of Industry and Information Technology publicly solicited opinions on the “Preparatory Plan for the Metaverse Standardization Working Group of the Ministry of Industry and Information Technology (Draft for Comment)”. Regarding the post-establishment work plan, the opinions stated:
Kuaishou’s self-developed Wenshengtu large model “Ketu” has recently started internal testing
According to a report by Future Science and Technology on September 18, Kuaishou’s latest progress in the field of AIGC, the self-developed Vincent graph model “Kolors”, has been fully launched within the company for internal testing. According to sources from the R&D members of the Ketu project team, the “Ketu” Wenshengtu large model has three outstanding features, namely strong text understanding, rich detail characterization, and diverse style transformations. As previously reported, Kuaishou has launched a new AIGC application scenario, and the AI game review function has started internal testing.
DeepMind co-founder: Generative AI is just a stage, the next step is interactive AI
According to DeepTech, DeepMind co-founder Mustafa Suleyman said in an interview last Friday that generative AI is just a stage in the development of artificial intelligence. “The next step is interactive AI.” , robots can accomplish the tasks you set for them by calling other software and other people.”
Alibaba Cloud Tongyi Industry Large Model Copyright Approved
According to Jiemian News on September 18, the China Copyright Protection Center website shows that on September 13, the software copyright of Hangzhou Alibaba Cloud Feitian Information Technology Co., Ltd.’s “Tongyi Industry Large Model Platform” was registered and approved, and the current version number is V1.0 . Hangzhou Alibaba Cloud Feitian Information Technology Co., Ltd. was established in March 2019. Its legal representative is Jiang Jiangwei, with a registered capital of 1 million yuan. It is wholly owned by Zhejiang Alibaba Cloud Computing Co., Ltd.
Baidu’s “Wenxin Yiyan Software” copyright has been registered and approved
According to a report from the Financial Associated Press on September 18, Tianyancha showed that the copyright of “Wenxin Yiyan Software” of Beijing Baidu Netxun Technology Co., Ltd. was recently approved for registration, and the current version number is V1.0.0. It is reported that “Wen Xin Yi Yan” is Baidu’s AI language model. Previously, Baidu had applied to register multiple “Wen Xin Yi Yan”, “Wen Xin Yi Yan ERNIE BOT” and graphic trademarks. As previously reported, Baidu released the Wen Xin Yi Yan plug-in ecological platform “Lingjing Matrix”.
Wayve, a startup company, releases LINGO-1, a large interactive autonomous driving model, which deeply integrates large models and autonomous driving.
According to a report by the Science and Technology Innovation Board Daily on September 18, the startup Wayve recently launched the large-scale autonomous driving interaction model LINGO-1, which deeply integrates the large language model and autonomous driving. After being trained on a variety of visual and language data, LINGO-1 can not only perform visual question answering (VQA) tasks such as perception, counterfactuals, planning, reasoning, and attention, but also describe driving behavior and reasoning. It is reported that Wayve is a British start-up company dedicated to realizing autonomous driving through reinforcement learning.
The Zhiji LS6 car is equipped with a large generative model of Zhiji and has functions such as “one-click AI driving”
According to a report by IT House on September 17, SAIC Motor recently accepted a survey from investors and stated that the new car Zhiji LS6, which was released at the Chengdu Auto Show at the end of August, has “one-click AI driving”, “one-click escape”, “one-click escape” and “one-click AI driving”. Innovative functions such as “edge tracking” and “one-click tracking”. It is reported that the car is equipped with a Zhiji generative large model (Beta version), which is based on a new multi-modal perception and AI algorithm. It is said to “integrate hundreds of billions of parameters + machine learning capabilities + natural language processing” and can accurately identify all instructions and has multiple functions. Human interaction. Previously reported, Zhiji Automobile Liu Tao: The industry’s first vertical GPT large model will be delivered and launched in October.
Microsoft announces early access program for M365 Copilot, introducing AI productivity tools to the first Australian organizations
According to a report by Webmaster Home on September 18, Microsoft announced that the first batch of Australian customers are eligible for the Microsoft365 (M365) Copilot early access program, which is a global invitation-only program. Australian electricity and gas company AGL, data solutions provider Data#3, healthcare company Bupa, National Australia Bank (NAB), electricity transmission company Powerlink Queensland, superannuation management company REST Super and insurance company Suncorp are among the first batch of Australian companies Mechanism using M365 Copilot.
It is reported that M365 Copilot is a powerful generative AI service that combines large language models (LLMs) with Microsoft365’s organizational data, enabling these enterprises to take advantage of the latest generative technologies in applications such as Microsoft Teams, Word, PowerPoint and Excel. AI technology to achieve new levels of productivity.
Google modifies its “Useful Content Updates” policy and admits the difficulties in supervising AI-generated content
According to a report by Webmaster Home on September 18, Google’s latest “Useful Content Update” has its third iteration and aims to change its stance on artificial intelligence-generated content. The update was first rolled out in August 2022, with the main goal of classifying content as “written for search engines” or “written for people.” Google has tweaked the wording in its official Useful Content reference, with previous versions emphasizing that content generated by its “Useful Content System” is original, useful content “written for humans.” However, the latest update puts the focus on “original, useful content created for people by the Useful Content System.” The change reflects Google’s growing recognition of the difficulty of regulating content generated by artificial intelligence, emphasizing that content should be human-centered, regardless of how it is created. Additionally, OpenAI removed its low-accuracy AI recognizer and plans to research more efficient text source technology to help users understand whether audio or visual content was generated by artificial intelligence.
Recommended reading for the Metaverse:
“AI Field: Recent Technology Highlights and Investment Opportunities”
Artificial intelligence technology is developing at a rapid pace, and there are many start-up companies in the field of artificial intelligence that stand out. In this article, let’s take a look at which promising AI startups have been picked up by investors from the likes of Sequoia Capital and Kleiner Perkins Caufield & Byers.
“First-hand | Kuaishou’s self-developed Wenshengtu large model “Ketu” is here, and has started internal testing in the near future”
Since the second half of this year, Kuaishou has been frequently developing large model business. GenAI New World learned first-hand that Kuaishou’s latest progress in the field of AIGC - the self-developed Vincent graph model “Kolors” has been fully launched within the company for internal testing. Following the launch of the large language model “KwaiYii” last month, Kuaishou launched the large Wensheng graph model “Kolors” in less than a month. "Afterwards, the territory of “Vincent Pictures” was added.
“Revealing the Transformer in iPhone: Based on GPT-2 architecture, word segmenter contains emoji, produced by MIT alumni”
The “secret” of Apple’s Transformer has been revealed by enthusiasts. A little brother named Jack Cook turned the macOS Sonoma beta upside down. As a result, he really dug up a lot of fresh information: In terms of model architecture, Brother Cook believes that Apple’s language model is more based on GPT- 2 built. In terms of tokenizers, emoticons are very prominent.