Jin10 data reported on August 26, early this morning, Microsoft Research open-sourced the innovative audio model VibeVoice-1.5B. VibeVoice-1.5B has achieved multiple significant technological breakthroughs in the field of voice: it can continuously synthesize 90 minutes of ultra-realistic voice in one go, while most previous models could only synthesize less than 60 minutes of voice and would face issues like tone drift and semantic disconnection after 30 minutes.

Google will launch the brand new AI voice assistant Gemini for Home in October, replacing Google Assistant, supporting more natural voice interactions and complex home control. Gemini Live will be released simultaneously, enhancing the multi-turn conversation experience, and may introduce new Nest hardware.

Odaily News X platform founder Musk stated that Grok will soon be able to speak, which means it will have voice output capabilities.

Jin10 reported on August 12 that Gaode has integrated the Tongyi large model. Based on the visual, speech, and text modalities of Tongyi Laboratory, Gaode has built multiple specialized models for multi-modal spatial perception, behavior cognition, and spatiotemporal intention understanding, and launched the world's first AI-native travel agent, which can assist users in formulating complex travel plans through voice interaction. Users can experience it by searching "spatial intelligence" in the latest version of the Gaode Maps APP.

ChainCatcher news, EdgeX Labs announced the launch of a brand new official website, fully presenting the hybrid AI + DePIN technology architecture that integrates cloud and edge, along with the global Computing Power network layout. This upgrade focuses on PaaS and SaaS level AI services, deeply integrating multimodal large language model capabilities, hybrid AI Computing Power scheduling protocols, and the global Node network, providing developers and enterprises with full-link support from model deployment, personalized fine-tuning to real-time inference. At the same time, the newly integrated TTS voice service and multimodal interaction system enable text, images, and voice to achieve low latency and efficient collaboration within the same intelligent network. EdgeX Labs Smart Hardware Matrix: PoS GPU-Server for enterprise-level high concurrency inference, PoW AI Dual

On July 31, Jin10 reported that trust must be upheld and actions must be resolute. It is not easy for the China-U.S. trade talks to reach Sweden, and the future path still requires both sides to move towards each other. Both sides should continue to follow the important consensus reached during the conversations between the two heads of state, fully leverage the mechanisms for China-U.S. trade consultations, and continue to implement the consensus and results that have already been formed. Efforts should be made to continuously enhance consensus, reduce misunderstandings, strengthen cooperation, and deepen dialogue and consultations, striving for more win-win outcomes, thereby injecting more positive energy into the stable development of the economies of both countries and the world.

Roam Premium eSIM has launched a new communication solution that supports calls, SMS, and high-speed data in over 160 countries worldwide. Users can activate it through the Roam App without the need for a physical SIM card, and they will be billed at local rates. The service supports various deposit methods, and in the future, payment with ROA Token will be added. Additionally, the Roam App has introduced an ad-watching feature, allowing users to earn extra rewards by watching advertisements.

Meta Platforms acquisition of voice artificial intelligence startup PlayAI to enhance its audio technology and AI initiatives. The PlayAI team will join Meta to support its AI roles and audio content development, aligning with the technology industry's trend towards conversational interfaces.

During the bankruptcy liquidation process of FTX, a motion proposed by the FTX estate sparked significant controversy. The motion requested to suspend payments to residents of jurisdictions with legal or regulatory restrictions related to crypto assets trading. However, a Chinese creditor, Weiwei Ji, objected, pointing out that despite residing in Singapore, they were classified as Chinese creditors due to holding Chinese passports and faced the risk of being excluded from compensation. This objection represents the voice of Weiwei Ji and an increasing number of Chinese creditors (according to documents, this group exceeds 300 individuals) who question the fairness and legality of this motion.

Gate News bot message, Musk announced through social media that if the "Great and Beautiful" spending bill is passed, he will start forming a new political party in the United States the next day. He emphasized: "Our country needs an alternative to the single-party Democratic Republican Party so that the people truly have a voice." According to Jinshi report, the U.S. Senate...

Jin10 data June 26th news, U.S. President Trump posted on social media: "Why are Republicans allowing the Democratic 'mouthpiece' Voice of America (VOA) to continue to exist? This is completely a leftist disaster – no Republican should vote to support its existence. Let it disappear!"

Gate News, the well-known crypto market maker Jump Crypto officially announced its return, reappeared as a "builder", and announced the results of the five-year construction. The company is positioned as a team of engineers at its core, and its strategy is to identify industry constraints and build solutions. Key achievements include: Pyth Network: A high-performance oracle with over 280 million daily updates, serving 100+ chains and 520+ applications. Wormhole: A cross-chain protocol that has delivered more than 1 billion messages, connecting the multi-chain ecosystem. Firedancer: Solana's high-performance standalone validator client designed to scale the network. DoubleZero: Build a decentralized, low-latency optical fiber network to solve the bottleneck of public network speed. Incubation security companies: Cordial Systems ( self-custody wallet ), Asymmetric Research ( embedded security, and has prevented losses of more than $5 billion ). All projects adhere to the principles of open source and decentralization, and Jump plays the role of a core contributor. At this time, the voice stems from the dual confidence in the technological development of the industry and the improvement of the policy environment (such as a more open attitude of the regulators). For the first time, Jump was deeply involved in U.S. regulatory discussions, submitting a policy advisory letter calling for a clear and rational crypto regulatory framework. At the end of the article, an invitation is issued to builders, ecological contributors and policy participants to build the future together.

PANews, June 19 news, OpenAI announced that the "Recording Mode" of ChatGPT has officially been launched for Pro, Enterprise, and Edu users, currently supporting macOS desktop applications. This feature can record meetings, brainstorming sessions, or voice notes, and automatically transcribe audio, extract key points, and generate follow-up action plans.

Golden Finance reports that Google: Issues with Gmail, Calendar, Chat, Cloud Search, Google Drive, Meetings, Tasks, and Google Voice have been resolved.

Gate News bot message, the average trading volume on the Bitcoin network over the past 7 days has decreased to 317,000 transactions, hitting a new low since October 2023. Data shows that only 256,000 transactions were recorded on the blockchain on June 1. In this context, some miners have begun to accept transactions below the default forwarding standard of 1 sat/vB. In response to the current situation, 31 Bitcoin Core developers jointly published an open letter on June 6, explicitly opposing nodes filtering low-fee or non-standard transactions. They pointed out that such practices violate Bitcoin's censorship resistance principles while emphasizing that private channels like Slipstream will have a negative impact on decentralization.

Jin10 Data, June 6th – Nick Timiraos, the "voice of the Federal Reserve", stated in a post that, based on unrounded data, the unemployment rate in May rose from 4.187% in April to 4.244%. The highest unemployment rate last year occurred in November 2024, at 4.231%. It can be said that the unemployment rate in the U.S. in May is the highest level since October 2021 (when it was 4.500%), based on unrounded data.

Jin10 reported on May 16 that Tencent Hunyuan today released the industry's first millisecond-level response real-time image generation model Hunyuan Image2.0 (parameter scale increased by 10 times), supporting various interaction methods such as text, voice, and sketches. In addition, Tencent Cloud also revealed that Tencent Hunyuan's native multimodal image generation model is already on the way.

Jin10 data May 13 news, the article points out that the significance of the China-U.S. Geneva economic and trade talks lies not only in the specific results achieved but also in the fact that both sides reaffirmed the path of equal dialogue and negotiation to resolve differences. The road ahead may not be smooth, but the Chinese side is willing to work with the U.S. side, taking this high-level economic and trade meeting as a good foundation, to continue with a pragmatic attitude towards problem-solving, maintain the momentum of dialogue, deeply explore cooperation potential, extend the cooperation list, enlarge the cooperation cake, and promote the healthy and stable development of China-U.S. economic and trade relations, benefiting the people of both countries and the world.

The high-level economic and trade talks between China and the United States will be held in Switzerland. The Chinese side is firm in safeguarding its own interests, advocates for win-win cooperation, calls for mutual respect from the US side, and seeks to resolve differences through equal dialogue, promoting the healthy development of China-US economic and trade relations and injecting certainty into the stability of the world economy.

Jin10 reported on May 5 that in March of this year, Microsoft (MSFT.O) announced that it will terminate the operation of Skype on May 5, 2025, and recommend users migrate to the Microsoft Teams application, marking the imminent end of this once-dominant internet voice and messaging service, concluding 14 years of operation.

PANews, April 29 news, Meta has launched the Meta AI application. Meta announced that the voice conversation feature (including full-duplex demonstration) in its Meta AI application has been launched in the United States, Canada, Australia, and New Zealand.

Jin10 data reported on April 26th, today, Kimi released a new Open Source project - the brand new general audio foundation model Kimi-Audio. According to the introduction, this model supports various tasks such as speech recognition, audio understanding, audio to text, and voice dialogue.

Golden Ten Data reported on April 17 that Fliggy launched a new AI product "Ask a Question", which integrates multiple main models of Tongyi Qianwen and supports multi-modal input, allowing users to interact with text, voice and even in dialects.

A former OpenAI employee submitted a legal brief today in support of Musk's lawsuit to maintain the company's non-profit status. The former employee stated that depriving the non-profit organization of control would severely violate the company's mission, and that overseeing for-profit entities is crucial to the company's strategy, as well as affecting recruitment and employee incentives.

PANews, April 10th news, according to TestingCatalog, xAI is preparing to launch the Grok voice mode with visual capabilities on iOS. Camera access is currently open, and a speech rate adjustment option will be added later. Meanwhile, xAI has launched grok-3-beta, grok-3-mini-beta, and their fast versions in the API. Users who deposit 5 dollars and participate in the plan can access these models with a monthly bonus of 150 dollars.

Golden Ten Data on March 16, according to US media reports, more than 1,300 employees of Voice of America Radio were forced to take leave on the 15th. The day before, U.S. President Donald Trump ordered a reduction in the size of the U.S. Agency for International Media, which is responsible for allocating funds to Voice of America radio.

Trump ordered cuts to U.S. government-funded media outlet Voice of America and six other organizations, resulting in the suspension of dozens of employees with full wages and benefits. The move is part of the administration's downsizing of the bureaucracy, which Trump signed on March 14.

Golden Ten Data on March 16, Feige Cloud Communication issued a statement on March 16, saying that the voice products involved in the "CCTV 3.15 Gala" report are a module of our SaaS products, which is only used as a tool for customers to use, we do not participate in the specific business operations of customers, and we have also made strict restrictions on customer qualifications and customer data sources, but this incident reflects our company's deficiencies in customer supervision and service management, and insufficient professional knowledge training for internal personnel. We are deeply remorseful for this and sincerely apologize to all users, partners and all sectors of society.

On March 15, U.S. President Donald Trump signed an executive order on the 14th aimed at weakening the parent agency of the U.S. government-funded media agency "Voice of America" and six other federal agencies, as the latest move by the government to reduce bureaucracy. These include one that provides financial support to museums and libraries and one that deals specifically with homelessness. The order requires these agencies to scale back their operations to the minimum required by law.

Jinshi data news on March 13th, Zhongguancun Kejin released DeZhu's large model platform 2.5 and DeZhu's intelligent customer service 4.0. Among them, DeZhu's large model platform 2.5 has now accessed DeepSeek, covering Computing Power, data, models, intelligent entities, and other full-chain large model development and application capabilities. DeZhu's intelligent customer service 4.0 also has the full-stack capability of intelligent customer service products, covering a full set of intelligent customer service products including full-media contact centers, voice robots, and text robots. Zhongguancun Kejin stated that it is expected that by 2025, over 90% of enterprises will integrate large model Depth into their core business.

BlockBeats news, on March 3rd, Ronaldinho X account posted, stating, 'We have heard the voice of the community and will continue to listen. 1. Ownership not relinquished? Now relinquished! Token cannot be destroyed without notice. 2. 30-day Lock-up Position not enough? Now Lock-up Position for 255 years. Security is not a detail, but a top priority. Your trust is crucial, we are here for long-term development.'

Jinshi data news on March 2nd, according to Agence France-Presse reported on March 1st, Ukrainian President Zelensky insisted on March 1st that the support of US President Trump is still "crucial" for Ukraine. Prior to this, he had an inappropriate argument with the US President, which led to urgent responses from Ukraine's European allies on the issue of Russia-Ukraine conflict. Zelensky posted on social platform X: "The support of President Trump is crucial for us. He wants to end the war, but no one desires peace more than us."

On February 26, Jinshi Data reported that following the LanTu Zhiyin becoming the first model in the industry to integrate DeepSeek capabilities, there was news recently that LanTu has another announcement. Its self-developed AI voice dialogue system will enter a position in the first half of the year, with a vehicle control response speed of less than 1 second and a wake-up recognition rate exceeding 98%. In the second half of the year, LanTu's models will also undergo a large-scale OTA update, with the new version further enhancing the vehicle AI functionality.

Odaily Planet Daily News, Musk announced on the X platform that a major update to the Grok voice mode will be launched tomorrow night (February 25th). (Jinshi)

Jinse Finance reported that Elon Musk announced on social media that Grok's voice mode, memory function, and many other improvements will be launched this weekend. Musk also mentioned in interaction with netizens that Grok-3 has been a complete success, and the XAI team has indeed done an excellent job. When asked about the most anticipated part, Musk clearly stated that it is the voice conversation function.

Odaily Planet Daily News, Musk said that Grok 3 voice mode may be launched in about a week. (Jinshi)

BlockBeats news, on February 18th, xAI engineers said that the Grok voice assistant is currently being improved and will be available to users in the coming weeks. (Jinshi)

Elon Musk and the xAI team announced the introduction of the AI search engine DeepSearch at the Grok 3 conference, described as an early version of intelligent physical ability. The voice mode may be launched next week, and engineers say DeepSearch is a versatile agent tool that can help write code and solve daily problems.

OpenAI CEO Sam Altman has announced the roadmap for GPT-4.5 and GPT-5. GPT-4.5 is the last non-'chain thinking' model, expected to be released in the coming weeks or months. After that, OpenAI will integrate the o series with the GPT series models to make them more intelligently adapt to different task requirements. GPT-5 will combine the o3 model and integrate voice, Canvas, search, and Depth research functions. ChatGPT users can use the standard intelligent version of GPT-5 for free, while Plus and Pro users can access higher levels of intelligence.

Golden Ten Data February 10th News, according to Jiangqi Group, recently, Jiangqi Group has integrated the DeepSeek large model into its Zhiling cockpit system, realizing more accurate intent understanding, voice interaction, proactive reminders, intelligent vehicle control, and other features. This means that Jiangqi Group's full-vehicle AI Zhiling assistant has taken another important step towards its goal. In addition, based on the powerful reasoning ability of the DeepSeek model, Jiangqi Group will also research its application in software code development, testing problem diagnosis and positioning, and automated testing to improve development efficiency.

Tomcat AI emotional companion robot is testing multiple model capabilities such as DeepSeek, carrying Tomcat emotional companion vertical model and West Lake Xincheng's voice synthesis model, and calling some models in the DouBao series. The company stated that it will continue to integrate different AI large models to enhance and supplement the product's functionality.

On January 23, Jin Shi Data reported that a memorandum showed that the US Department of Justice has ordered the Civil Rights Department to stop the litigation that the Biden administration is conducting, and not to initiate any new cases or settlements. The memorandum did not specify how long the freeze would last. The memorandum pointed out that the implementation of the freeze is to ensure that the federal government "makes a voice in legal opinions" and ensure that the President-appointed or designated person has the opportunity to decide whether to initiate any new cases. It is reported that Harmeet Dhillon, the head of the Civil Rights Division of the Justice Department nominated by Trump, is responsible for this matter.

On January 22nd, Jinshi Data reported that the first end-to-end speech model GLM-4-Voice by Zhipu AI has officially launched on the open platform. It can directly understand and generate Chinese and English speech, realize real-time speech conversation, and flexibly adjust the emotions, tones, speeds, dialects, and other characteristics of speech according to user instructions, making voice interaction more natural and vivid.

Dou Bao launches a lifelike voice conversation AI model, which has higher naturalness and emotional richness compared to GPT-4o, and is expected to break through the ability of lifelike voice conversation, giving AI conversation a 'human feeling'. It is beneficial to further enhance the user experience of emotional companion applications and the development of AI-side hardware. It is recommended to continue to follow investment opportunities related to AI companionship.

Vitalik suggests providing anonymous broadcasting tools for ETH holders, using ZKP to confirm the holding amount, and allowing one message to be posted every (1.2 billion/N) seconds to avoid social media volume distortion, express the voice of the ETH holder community, and hope that related platforms will establish a Smart Contract-friendly logic to implement the entire mechanism.