<div>
  <p>
  </p>
  <ul>
    <li>
      <p>
        Alibaba Cloud has launched &ldquo;Qwen2-VL,&rdquo; an advanced vision-language AI model to enhance visual understanding, video comprehension, and multilingual text-image processing.
      </p>
    </li>
    <li>
      <p>
        The model is available in 72 billion, 7 billion, and 2 billion parameters. Additionally, it can analyze and describe handwriting in multiple languages, identify objects in images, analyze live video in near real-time, and analyze videos longer than 20 minutes while answering questions about content. Furthermore, it supports English, Chinese, most European languages, and Japanese, Korean, Arabic, and Vietnamese.&nbsp;
      </p>
    </li>
    <li>
      <p>
        The company claims the model can be integrated into mobile phones and robots, allowing automated operations based on visual environments and text instructions. It also supports function calling, enabling integration with third-party software and visual extraction of information from external sources.
      </p>
    </li>
  </ul>
  <p>
  </p>
</div>


<div>
  <p>
    <a href="https://venturebeat.com/ai/alibaba-releases-new-ai-model-qwen2-vl-that-can-analyze-videos-more-than-20-minutes-long/">
      VentureBeat
    </a>
  </p>
  <p>
    <a href="https://www.aibase.com/news/11386">
      Aibase
    </a>
  </p>
  <p>
  </p>
</div>


Foundation Models

INCUMBENTS_PRESENT

PROMINENT_CATEGORY

<div>
 
 <p>Alibaba Cloud presents an innovative vision-language AI model "Qwen2-VL." This advanced model is designed to improve visual understanding, multilingual text-image processing, and video analysis.</p>
 <p>The parameters of the model are available in 72 billion, 7 billion, and 2 billion. This AI model can process handwriting in different languages, recognize objects in images, and perform live video analysis nearly in real time. The model can handle videos that extend beyond 20 minutes and provide answers related to the content. Apart from English and Chinese, it also supports Thai, Japanese, Korean, Arabic, Vietnamese, and a majority of European languages.</p>
 <p>Alibaba says that this model can be integrated into mobile devices and robots for automation based on visual environments and text instructions. Additionally, it supports function calls which allows for integration with third-party software and visually extracting information from external sources.</p>

</div>

Alibaba releases AI model ‘Qwen2-VL’ for video and image analysis

Exploring the limitless possibilities of AI

Generative AI Applications

<div>
 <p>T-Mobile, a wireless provider, is working with OpenAI to develop IntentCX, an AI-based customer service platform. They aim to combine T-Mobile's customer relationship experience with OpenAI's AI technology to deliver personalized service, genuine engagement, and proactive actions. With the customer's permission, the platform will strive to understand client requirements, present solutions, and take actions on their behalf.</p>
 <p>T-Mobile aims to integrate IntentCX into its operations by 2025 and plans further collaborations with OpenAI to create additional AI-enabled tools and services. The platform is already being tested by T-Mobile. It has access to T-Mobile's data and is receiving training from the company's customer service team.</p>
</div>

T-Mobile partners with OpenAI to develop AI-powered customer service platform

<div>
 <p>AI video generation startup, Runway, has forged a partnership with major Hollywood studio, Lionsgate, to train a personalised video model based on Lionsgate's extensive film catalog.</p>
 <p>This alliance will provide Lionsgate's filmmakers, directors, and creative professionals access to the video model in order to improve their work. Simultaneously, Runway has plans to look into licensing models as templates, enabling individual creators to construct and train their own custom models.</p>
 <p>This collaboration represents the first known publicly announced partnership between a Generation AI company and a distinguished Hollywood studio.</p>
</div>

Runway partners with Lionsgate to develop AI video tools using studio's movie catalog

Quantum information processing is slowly becoming a reality, ushering in a new technological era.

Quantum Computing

<div>
 
 <p>QMill, a Finland-based company, has successfully raised EUR 4 million (equivalent to USD 4.6 million) in a seed funding. Maki.vc and Antler led this funding round, alongside Kvanted and Business Finland also making contributions.</p>
 <p>The funds raised in this round will be used to fast-track QMill's initiative to apply practical quantum computing solutions to industrial uses. The plan includes enhancing research and development efforts as well as expanding the team. The company aims to employ 20 additional staff members as part of this plan.</p>
 <p>QMill specializes in the development of efficient quantum algorithms for solving complicated optimization problems prevalent in multiple industries. Using near-term quantum technology, QMill can deliver favorable outcomes for real-world industrial applications. Additionally, the company is looking into optimizing LLMs and contributing to AI developments.</p>
  
</div>

QMill raises EUR 4 million in seed funding to provide quantum computing industrial applications

<div>
 
 <p>QuiX Quantum has launched the Bia Quantum Cloud Computing Service, marking its entry into the quantum-as-a-service (QaaS) market. The new service is designed to give businesses and research organizations convenient access to quantum computing resources, blending quantum and traditional computing spaces.</p>
 <p>The Bia system comes with a single photon generator and light detector modules along with quantum control software. It is designed to handle specific challenges in basic research and quantum simulations, encompassing Boson Sampling and Quantum Monte Carlo simulations. The system utilizes a variational quantum eigensolver (VQE) algorithm to address optimization problems when merging traditional computing resources with its quantum hardware.</p>
 
</div>

QuiX Quantum launches 'Bia' quantum cloud computing service for quantum solutions

<div>
 
 <p>Oxford Ionics, a trapped-ion quantum computing startup, and Infineon Technologies, a semiconductor company, have joined forces to create MinIon, a portable quantum computer for Cyberagentur, Germany's cybersecurity agency. </p>
 <p>This collaboration utilizes Oxford Ionics' e-qubit control tech and Infineon's microfabrication expertise to create a high-performance, minimalistic computer. MinIon is engineered to be used in mobile defense scenarios and locations lacking stationary data centers. With the integration of high-fidelity qubits, the system is expected to enhance both computational power and encryption abilities for national security tasks.</p>
 
</div>

Oxford Ionics and Infineon Technologies partner to build portable quantum computer for Cyberagentur

<div>
 <p>Tencent AI Lab, in collaboration with scientists from Johns Hopkins University has released a new text-to-audio generation model named EzAudio. The aim of EzAudio is to manufacture high-quality sound effects using just text inputs. For the sake of fostering further research, Tencent has made the model code, dataset, and checkpoints publicly available.</p>
 <p>In order to enhance its performance and efficiency, several technical enhancements have been incorporated into the model. One of these is a fresh adaptive layer normalization technique called "AdaLN-SOLA," others are long-skip connections and positioning techniques namely RoPE (Rotary Position Embedding).</p>
 <p>As per the researchers, EzAudio surpasses present open-source models in both objective and subjective evaluations. The model has the potential to be utilized in a variety of fields, such as voice and music production, entertainment, media, virtual and accessibility assistants, and services.</p>
</div>

Tencent Ai Lab launches EzAudio AI for text-to-audio generation with Johns Hopkins University

A decentralized version of the web powered by the blockchain

Web3 Ecosystem

<div>
 
 <p>TON, a blockchain network, received a USD 30 million investment from crypto exchange Bitget and Web3 investment firm Foresight Ventures through a token exchange. The investment aims to support TON projects, particularly in the 'tap-to-earn' gaming field integrated with Telegram. Both Bitget and Foresight Ventures will participate in shaping TON's governance. TON is among the rapidly growing blockchains in 2024, with its total value locked jumping 18-fold to USD 350 million within the last six months. The TON ecosystem houses over 1,000 dApps, serving millions of users.</p>
 
</div>

TON secures USD 30 million in investment from Bitget and Foresight Ventures

<div>

<p>Hemi Labs, which develops blockchain infrastructure, has raised USD 15 million in a funding round. Major investors include Binance Labs, Breyer Capital, and Big Brain Holdings, along with Crypto.com, Web3 Ventures, HyperChain Capital, Alchemy, and other investors.</p>
<p>The raised funds will be channeled towards the development and launch of the Hemi Network. This is a modular blockchain network based on Bitcoin and Ethereum, with a target launch date in Q4 2024. The network's incentivized test net is already operational.</p>
<p>The primary objective of Hemi Network is to integrate Bitcoin and Ethereum into a singular supernetwork. This network will offer a variety of features, such as Hemi Virtual Machine, Bitcoin programmability, Superfinality, trustless cross-chain portability, and asset programmability while ensuring superior scaling, security, and interoperability.</p>
<p>Established in 2024, Hemi Labs is a blockchain technology firm. It has developed the Hemi VM (hVM) - a system that amalgamates a full Bitcoin node with an Ethereum Virtual Machine (EVM). The combination allows Hemi to create functionalities that link the Bitcoin and Ethereum networks. The company's key offerings contain security features leveraging Proof-of-Proof technology, asset transfer capabilities between Bitcoin and Ethereum through Hemi's Tunnels, and toolkits to develop Bitcoin-aware, EVM-compatible applications via its Hemi Bitcoin Kit (hBK).</p>

</div>

Hemi Labs raises USD 15 million in funding to launch blockchain network

The toolkit to build machine learning applications at scale

Machine Learning Infrastructure

<div>
 
 <p>Fivetran has launched "Hybrid Deployment," a solution for running data pipelines within a customer's environment that can be managed via Fivetran's platform. The data integration service provider aims to help businesses centralize data from different sources safely through this new solution.</p>
 <p>Hybrid Deployment comes with several features, such as a local connector for secure data transmission, a user-friendly interface for setup and monitoring, and an independent control and data plane structure to keep data within the customer's secure infrastructure.</p>
 <p>According to Fivetran, this new offering provides various advantages, including a comprehensive view of all pipelines from one platform, improved data protection and compliance, compatibility with various cloud and on-premises setups, easy installation and maintenance, the ability to scale and personalize pipelines, and comprehensive cost management tools.</p>
 
</div>

Fivetran launches Hybrid Deployment for data pipeline management

Data Infrastructure & Analytics

<div>

 <p>Fivetran has introduced "Hybrid Deployment," a new data integration solution. The product allows customers to operate data pipelines within their own environment while management is conducted through Fivetran's platform. This new solution aids enterprises in centralizing data from different sources securely.</p>
 <p>Hybrid Deployment features a local connector for secure data transfer, a user-friendly interface, and separate control and data plane architecture. This ensures that data stays in the customer's own secure environment.</p>
 <p>The company claims that Hybrid Deployment provides benefits such as full pipeline visibility, enhanced data security, compatibility with various environments, easy setup and maintenance, flexible scaling and customization of pipelines, and detailed cost management tools.</p>

</div>

All Updates

Alibaba releases AI model ‘Qwen2-VL’ for video and image analysis

T-Mobile partners with OpenAI to develop AI-powered customer service platform

Runway partners with Lionsgate to develop AI video tools using studio's movie catalog

QMill raises EUR 4 million in seed funding to provide quantum computing industrial applications

QuiX Quantum launches 'Bia' quantum cloud computing service for quantum solutions

Oxford Ionics and Infineon Technologies partner to build portable quantum computer for Cyberagentur

Tencent Ai Lab launches EzAudio AI for text-to-audio generation with Johns Hopkins University

TON secures USD 30 million in investment from Bitget and Foresight Ventures

Hemi Labs raises USD 15 million in funding to launch blockchain network

Fivetran launches Hybrid Deployment for data pipeline management

Fivetran launches Hybrid Deployment for data pipeline management

Foundation Models

Alibaba releases AI model ‘Qwen2-VL’ for video and image analysis

Contact us