Recently, Baidu Cloud launched a new AI feature that can turn a photo into an anime character. Users can open the Baidu Cloud app, search for "comic" and upload a photo to generate stylized character photos with just one click. It is reported that this feature can create "iconic scenes" effects from real life to various classic animations, meeting the personalized needs of users of different ages. With the exclusive deployment of workflow and optimized models after multiple lightweight processes, Baidu Cloud is able to process and generate character photos twice as fast as similar AI products on the market, taking only about 10 seconds.

PANews May 30th news, according to the official tweet from OpenAI, all ChatGPT free users can now use browsing, visual, data analysis, file upload, and GPTs functions. Previously, on May 14th, OpenAI announced the free access to its new flagship model GPT-4o and other features, including browsing, data analysis, and memory functions, but with usage restrictions.

#NewUserGuide 🪪 How do I authenticate on the Gate.io App? ➡️ Click the user icon in the top left corner ➡️ Click on "Authentication" ➡️ Fill in the basic personal information ✅ Upload and finish 👉 Go to certification now: https://www.gate.io/myaccount/kyc_home

#NewUserGuide 🪪 How to Complete KYC Verification on Gate.io app? ➡️ Find the user icon on Gate.io app ➡️ Go to "KYC (Identity Verification)" ➡️ Input basic information as required ✅ Upload & Complete 👉Verify now: https://www.gate.io/myaccount/kyc_home

Google's parent company, Alphabet Inc., has filed a lawsuit against two Chinese, accusing them of using the company's platform to develop fraudulent cryptocurrency apps that have been downloaded more than 100,000 times. Alphabet Inc. alleges that scammers used its platforms, Google Play and YouTube, to upload and promote fraudulent crypto applications, and that they were engaged in extortion after committing hundreds of wire transfer frauds. Bloomberg reported on April 4 that the lawsuit will be filed in federal court in New York on Thursday morning local time. （Cointelgraph）

Spanish fintech company MONEI has started a pilot of the euro Stable Coin EURM under the supervision of the Spanish Central Bank. According to an announcement published by MONEI on its website on January 29, the regulatory sandbox pilot was launched in January and is overseen by the Central Bank of Spain. The pilot will test the trading capabilities of EURM, where users will verify their identity, upload their Wallet, deposit up to 10 EUR into their account and exchange it for 10 EURM. （Cointelegraph）

AI startup Artisse AI recently closed a $6.7 million seed round led by The London Fund, as reported by Techcrunch on January 24. Artisse is now available as an AI image app, Artisse. The app allows users to upload photos of themselves to train the AI on their images, and then use text or image prompts to generate new photos of themselves in a variety of settings, poses, and styles. The company claims that the app has been downloaded more than 200,000 times to date, and its images have been viewed on social media by an estimated 43 million people. Its website states that users have created more than 5 million photos so far.

On November 22, it was reported that AI company Anthropic released a large language model (LLM) Claude 2.1, providing a 200K context window, which exceeds the 128K context of GPT-4 Turbo recently announced by OpenAI. Users can send Claude about 150K words or more than 500 pages of information, meaning they can upload an entire codebase, financial statements, or long-form literary work for Claude to summarize, perform Q&A, predict trends, compare and contrast multiple documents, and more. In addition, Claude 2.1 includes a new tool usage feature that allows the model to be integrated with the user's existing processes, products, and APIs.

PANews reported on November 21 that Poly Network, a cross-chain interoperability protocol, disclosed details of the security incident investigation in early July, when the vulnerability that led to the incident affected 58 assets on 11 blockchains. The analysis found that the attacker implanted a Trojan virus in the program compilation environment, so as to obtain the consensus key of the Poly Network relay chain. Subsequently, they use these keys to forge cross-chain transactions. Hackers implant Trojan blocks of code during program compilation, and obtain and upload consensus keys when the program starts. They then use these keys to sign the block headers of the fake Poly Network Relay Chain, and eventually submit the fake cross-chain transactions and block headers to the target chain to execute the cross-chain exploit.

According to a report by Tech Planet on November 13, Zhang Yueguang, the head of the Miao Ya camera product under Ali Da Entertainment, has left his post, and he has participated in the planning of Alipay's Spring Festival "Jiwufu" and "Xiuyixi" projects, and has worked in Byte, Ali and other companies. Miao Ya camera is an AIGC product that is out of the circle at the C-end level, as long as you upload more than 20 photos containing faces, pay 9.9 yuan, and then choose the template and shape you like, you can make your own "digital clone", so as to get a photo work. Miao Ya camera "dominated" the list of application products for a period of time, but the subsequent ranking fell all the way. As of November 13, the latest data from Qimai shows that Miao Ya Camera's iOS "social" list ranks 64th. At present, how to form a long-term demand for AIGC products has become a difficult problem at the market level.

PANews reported on November 7 that, according to official news, Berklee College of Music announced a partnership with the NEAR Foundation to help build a music licensing platform. The parties will leverage the NEAR protocol's easy-to-use technology stack to create a marketplace where musicians can upload their music and creators can license music from Berklee College of Music students. THE PROJECT WILL BE RELEASED UNDER AN APP CALLED RADIAR, AND THANKS TO NEAR'S Web3 TECHNOLOGY, ALL TRANSACTIONS WILL BE RECORDED ON THE NEAR BLOCKCHAIN, ENSURING THAT CREATORS' RIGHTS AND REPRESENTATION ARE RECORDED ON-CHAIN. RADAR will use smart contracts to eliminate the licensing process and ensure that musicians receive the full economic value and prestige of their work.

According to the Securities Times E Company, the China Securities Construction Investment Research Report said that the first Open AI developer conference will be held on November 6. We expect the following updates: 1. Launch of an API with in-memory storage to reduce repeated uploads of history, increase the computation speed of large models, and reduce the cost of API calls by 95%; 2. Continuous multi-modal layout: At the end of October, the GPT 4 grayscale test was launched, and two major functions were launched: "support for multi-format file upload" and "one-stop tool integration" to optimize the multi-modal experience. APIs for GPT-4V and DALL・E3 may be available. In the long run, Open AI is expected to promote the gradual iteration of large models into AI agents and build a new ecosystem based on ChatGPT large models. According to reports today, OpenAI will hold its first developer conference next Monday and will launch new products.

PANews reported on November 3 that according to Cointelegraph, a new malware "KandyKorn" found on Apple's macOS system is related to the North Korean hacker group Lazarus, which targets blockchain engineers on crypto trading platforms. According to Elastic Security Labs, "KandyKorn" is an invisible backdoor capable of data retrieval, directory listing, file upload/download, secure deletion, process termination, and command execution. Initially, the attackers spread Python-based modules through Discord channels by impersonating community members. The social engineering attack tricked community members into downloading a malicious ZIP archive called 'Cross-platform Bridges .zip that mimicked arbitrage bots designed for automated profitability. However, the file imports 13 malicious modules that work together to steal and manipulate information.

OpenAI is currently rolling out a new beta feature for ChatGPT Plus members, and there has been user feedback that this update includes the ability to upload and process files, as well as multimodal support, as reported by IT Home on October 30. This suggests that instead of selecting a mode like Browse with Bing from GPT-4's drop-down menu, the user will automatically guess what the user wants to do based on context. These new features give ChatGPT Plus members access to some of the office features offered by the ChatGPT Enterprise plan. It is reported that after the file is passed to ChatGPT, it takes a few seconds to digest the file and then it can process the file, such as summarizing the data according to prompts, answering questions, or generating data visualizations.

zPass launches the Aleo chain, a protocol designed to use zero-knowledge for age verification, allowing users to upload identity documents offline to private servers and share anonymous proof with organizations. John Reynolds, product manager at Aleo, said: "We wanted to create a child-friendly space that would avoid their exposure to potentially inappropriate content, so our main goal at the moment is towards age verification. ”（BlockWorks）

PANews news on October 12, Baidu Vice President Shang Guobin revealed yesterday that Baidu is about to launch a new AI native map, which fully applies the Wenxin model to reconstruct the product interaction of the map and bring users a new experience. Users can use natural language dialogue to wake up various functions and services buried in the map in one step, further improving the efficiency of travel and decision-making. According to reports, the upcoming version of Baidu Map V19 will be comprehensively reconstructed from the two dimensions of "new interaction" and "new navigation". In addition, this "AI wizard" also has a personified digital human image, in addition to the default image recommended by the system, users can upload pictures to generate their own exclusive digital human. It is reported that "Baidu World 2023" will be held in Beijing Shougang Park on October 17. At that time, Baidu founder, chairman and CEO Robin Li will bring a series of reconstructed AI native applications to a new appearance.

According to the Shanghai Securities News, Shang Guobin, vice president of Baidu, revealed on October 11 that Baidu is about to launch a new AI native map, which comprehensively applies the Wenxin model to reconstruct the product interaction of the map and bring users a new experience. Users can use natural language dialogue to wake up various functions and services buried in the map in one step, further improving the efficiency of travel and decision-making. It is revealed that Baidu Map V19, which is reconstructed based on the Wenxin large model, will launch a new "AI Guide". It has multiple rounds of natural language interaction capabilities, which is closer to the "human communication" experience. In addition, this "AI wizard" also has a personalized digital human image, in addition to the default image recommended by the system, users can upload pictures to generate their own exclusive digital human.

According to a Techcrunch report on October 7, the AI application EPIK has soared to the top of the App Store. It's a photo editing app that lets users generate a nostalgic, '90s-style "yearbook" photo of themselves as one of its many templates. Similar to other recent popular AI applications, EPIK works by having users first upload a series of selfies, which EPIK then uses to generate retro yearbook photos in which the user has different poses, different looks, and hairstyles. In the US App Store, EPIK is currently ranked No. 1, but its popularity is not high on Google Play, ranking No. 37. Since its debut in August 2021, EPIK has had 92.3 million total installs, including 4.7 million downloads in the United States, according to market intelligence firm Apptopia. But the company said the app started gaining traction on September 19 and soared further 10 days later.

Golden Finance reported that at the 2023 Shanghai Blockchain International Week and the 9th Blockchain Global Summit, Lan Zhenzhong, founder of West Lake Xinchen and professor of West Lake University, said that I think large models and Web3 are related to productivity and production relations. Large models have greatly improved productivity levels, but if they are to work well, they must be matched by corresponding production relations. I summarized that there are several problems in the implementation of large models: First, the cost of training is very high. Start-up companies have no incentive to open source their models. Models that cost millions of dollars to train are open sourced, but then they have nothing to do with me. It is difficult for them to open source them. But open source is very important for models. Many models now are black boxes. Many research institutions cannot afford to train their own models. If everyone is training, then everyone is reinventing the wheel. So open source is very important, but it needs corresponding incentives. mechanism. Second, the cost of reasoning is high. The current reasoning cost of a single conversation in GPT4 is 60 cents, which is much more expensive than my speech. The cost of reasoning is very high and it is very difficult to implement. GPT4 can be used in many places, but the cost is unaffordable. The third is data sensitivity. Samsung’s data was leaked by OpenAI before and caused a lot of uproar. The data we upload to the large model now is sensitive data. Many companies are unwilling to upload their own data. How to deal with these problems? I hope Web3 can help us solve these problems.

According to PANews on August 29, according to Blockworks, Marcus Feistl, chief operating officer of Limewire, an NFT market in the field of art and music entertainment, said that later in the third quarter, users will be able to use melodies, beats, and song libraries to create artificial intelligence-generated music. The content initially came from partner platforms and LimeWire-owned content before allowing musicians to upload their own content. All content created in the studio will be minted on the Polygon blockchain. The platform will split revenue between artists who upload original content and creators of new AI-generated content. Specifically, artists earn a portion of the advertising revenue generated by LimeWire (paid in native LMWR tokens) based on the number of views their content generates. The content can also be traded on the LimeWire secondary market as an NFT. Creators will receive royalties on this volume, currently paid in USDC. Creators can also limit access to their content by requiring fans to pay a monthly fee (denominated in USDC). While each user gets a certain number of free creations per month, those who use it more often will be billed, Feistl noted. Earlier in the news in May, NFT marketplace Limewire announced that it had raised more than $16 million through the LMWR token sale.

Jinse Finance reported that Marcus Feistl, chief operating officer of NFT market LimeWire, said that later in the third quarter, users will be able to use melodies, beats, and song libraries to create artificial intelligence-generated music. The content initially came from partner platforms and LimeWire-owned content before allowing musicians to upload their own content. All content created in the studio will be minted on the Polygon blockchain. The platform splits revenue between uploading artists of original content and creators of new AI-generated content. Specifically, artists earn a portion of the advertising revenue generated by LimeWire (paid in native LMWR tokens) based on the number of views their content generates. These contents can also be traded on the LimeWire secondary market as NFTs. Creators will receive royalties on this volume, currently paid in USDC.

According to PANews news on August 15, according to CoinDesk, "Avalanche" author Neal Stephenson (Neal Stephenson) and cryptocurrency pioneer Peter Vessenes collaborated on the metaverse blockchain Lamina1 to launch Betanet, as well as for Web3 builders and developers. The new Hub provided. During the Betanet, partners, world builders, validators, and developers can try out the network and tinker with their toolkits. The goal of the testnet, according to Lamina1, is to allow builders to create modules for the open metaverse “in a more durable and stable environment” before the upcoming mainnet launch. Betanet is an "evolution" of its testnet (Testnet), which improves Lamina1's user experience by simplifying the onboarding process and stripping away cumbersome language and processes. The platform now has an easier authentication process to help users create new blockchain wallets without having to keep their private keys, while the Hub acts as a dashboard for user profiles, transaction history, and as a discovery platform based on Lamina1 Tools for new experiences. Users will also soon be able to send, receive and upload digital items, such as music and 3D items, and can use these items to build on-chain experiences in a more stable and realistic environment. Lamina1 will roll out new features over the next 6 months, such as L1 name service, contact saving and management features, dedicated subnetworks, decentralized storage of digital items, and digital identity customization. Lamina1 has also taken steps to ensure its infrastructure is carbon negative by delegating a proof-of-stake consensus mechanism and building a process for new nodes requiring them to purchase and withdraw carbon credits before validating any subnet.