Category: AI Related

  • Artificial intelligence startup Perplexity launched its AI-powered web browser, “Comet”

    Perplexity AI has launched its first AI-powered web browser called Comet on July 9, 2025. Comet is designed to integrate advanced artificial intelligence capabilities directly into the browsing experience, allowing users to interact with enterprise applications like Slack and ask complex questions via voice or text. The browser features Perplexity’s AI search engine, which provides instant answers, real-time summaries with sources, and can assist with tasks such as summarizing emails, organizing tabs, and managing calendars.

    Currently, Comet is available exclusively to Perplexity Max subscribers, who pay $200 per month, with access initially limited to invite-only users and a waitlist. Perplexity plans to gradually expand access over the summer of 2025. The company emphasizes continuous feature development and improvements based on user feedback, aiming to create a smarter alternative to traditional browsers like Chrome and Safari by offering an AI assistant that actively helps users rather than just searching.

    This launch positions Perplexity as a notable player competing with major tech companies in the consumer internet and AI space, leveraging its expertise in AI-powered search and productivity tools to enhance web browsing

  • YouTube Monetization Policy Update 2025. What about the contents by created by AI?

    Starting July 15, 2025, YouTube will enforce stricter monetization rules targeting mass-produced, repetitive, and low-quality content, with a particular focus on AI-generated videos that lack meaningful human input. This update is part of YouTube’s effort to improve content originality and ensure that monetized videos provide genuine value, whether educational, entertaining, or informative.

    Let’s have a look at the key points of the new policy include:

    • Channels relying heavily on reused, repetitive, or minimally edited content—such as reaction videos, compilations, AI-generated commentary, or synthetic voice videos—risk losing monetization entirely, not just on individual videos.

    • To remain eligible for the YouTube Partner Program (YPP), creators must add clear value, commentary, or significant editing to reused or AI-generated content.

    • Fully AI-generated content with no human contribution will generally not be monetized.

    • Channels must meet existing thresholds (1,000 subscribers and 4,000 valid watch hours in the past 12 months or 10 million Shorts views in the last 90 days) but must also comply with the new originality standards.

    • The policy aims to discourage “copy-paste” style channels and clickbait-heavy uploads, promoting authentic voices and meaningful content.

    • YouTube has not yet detailed specific penalties but warns that channels failing to meet these standards could be demonetized or removed from the Partner Program.

    This update signals YouTube’s commitment to combating low-effort, automated content flooding the platform, especially from AI tools, and encourages creators to produce original, engaging, and thoughtfully crafted videos to maintain monetization privileges

  • Apple Working on ChatGPT-Style AI Assistant for Customer Support

    Aaron Perris, a contributor for MacRumors, discovered evidence in Apple’s code indicating that Apple is developing a new AI-powered feature called “Support Assistant” for the Apple Support app. This assistant is designed to provide generative AI-based customer support through chat interactions, allowing users to describe their issues and receive AI-generated solutions before being connected to a live Apple support agent if needed.

    Let’s have a look at the key details about the Apple Support Assistant found by Aaron Perris include:

    • It uses generative AI models to answer queries related to Apple products and services.

    • Users will be able to interact with the assistant via chat within the Apple Support app.

    • The assistant may support uploading images, PDFs, or documents to help diagnose problems.

    • Apple warns that the AI might produce incorrect or misleading responses, so users should not rely solely on it for critical advice.

    • This feature is not yet live but is expected to launch in a future update of the Apple Support app.

    • The development aligns with Apple’s broader AI initiatives, including partnerships with OpenAI and improvements to Siri under AI chief John Giannandrea.

    Aaron Perris’s findings highlight Apple’s move to integrate a ChatGPT-style AI assistant into its customer support ecosystem, aiming to enhance user experience by providing quick AI-generated help while maintaining the option to escalate to human advisors.

  • EssilorLuxottica and Meta Announce Long Term Partnership, Shaping the Future of the Smart Eyewear Category

    Meta Platforms has acquired nearly a 3% stake in EssilorLuxottica, the world’s largest eyewear maker, in a deal valued at approximately €3 billion (about $3.5 billion). This strategic investment deepens Meta’s commitment to the AI-powered smart glasses market and wearable technology sector.

    Let’s have a look at the key points about this investment and partnership:

    • Meta and EssilorLuxottica have a longstanding collaboration since 2019, producing successful smart glasses under the Ray-Ban Meta brand, which have sold over 2 million units as of early 2025.

    • The companies recently launched the Oakley Meta HSTN smart glasses, featuring AI integration, hands-free cameras, audio speakers, water resistance, and Meta AI functionalities.

    • EssilorLuxottica plans to increase smart glasses production capacity to 10 million units by the end of 2026.

    • Meta is reportedly considering increasing its stake to around 5% over time, signaling a long-term strategic interest.

    • This investment follows a broader industry trend, with competitors like Google also investing in smart eyewear companies.

    • The partnership aims to develop multi-generational smart eyewear products combining EssilorLuxottica’s expertise in eyewear and retail with Meta’s advances in AI and technology.

    This move highlights Meta’s growing focus on integrating artificial intelligence into consumer wearables and expanding its hardware ecosystem beyond VR and social platforms.

  • Hugging Face has officially launched the Reachy Mini

    Hugging Face has officially launched the Reachy Mini, an open-source desktop robot priced starting at $299 for the Lite version, with a fully autonomous Wireless version available at $449. The Reachy Mini is an 11-inch tall, expressive humanoid robot designed for AI experimentation, human-robot interaction, and creative coding. It features motorized head and body rotation, animated antennas, a camera, microphones, and speakers, enabling rich multimodal interaction.

    Key points about Reachy Mini:

    • Two versions:

      • Lite version ($299) requires an external computer for operation.

      • Wireless version ($449) includes onboard computing (Raspberry Pi 5), battery, and wireless connectivity.

    • Open-source design sold as a DIY kit aimed at developers, researchers, educators, and AI enthusiasts.

    • Programmable primarily in Python, with JavaScript and Scratch support coming soon.

    • Integrates with the Hugging Face Hub, giving access to thousands of AI models and allowing users to share robotics applications.

    • Shipping is expected to start in August 2025, with some assembly required.

    • The robot supports over 15 pre-loaded behaviors and a simulation SDK for development before receiving the physical unit.

    This launch marks a significant step in democratizing robotics by offering an affordable, accessible platform for AI and robotics development, contrasting with traditional robots that can cost tens of thousands of dollars.

  • Kinetics and Mitsui O.S.K. Lines Sign MOU to Develop World’s First Integrated Floating Data Center Platform

    Mitsui O.S.K. Lines (MOL) and Karpowership, through its energy transition initiative Kinetics, have signed a Memorandum of Understanding (MOU) to jointly develop the world’s first integrated next-generation floating data center platform. This collaboration was announced on July 7, 2025.

    Let’s have a look the key details of the partnership include:

    • The floating data center will be hosted on a retrofitted vessel, designed to provide scalable, mobile, and rapidly deployable digital infrastructure.

    • It will be powered by uninterrupted, dedicated energy from various flexible sources, including Karpowership’s Powerships, land-based grids, onshore solar farms, offshore wind, or other energy generation solutions depending on location and client needs.

    • Karpowership brings over 20 years of experience and more than 10,000 MW of installed floating power capacity across 19 countries.

    • MOL contributes its extensive maritime and offshore logistics expertise to the project.

    • The floating platform aims to overcome challenges faced by traditional land-based data centers such as power constraints, land scarcity, and permitting delays.

    This initiative builds on a prior collaboration between MOL and Karpowership from 2019, where they cooperated on the LNG-to-Powership business under the brand “KARMOL,” focusing on providing LNG-fueled floating power solutions with environmental benefits and cost efficiency.

    The latest MOU marks a significant step toward combining maritime logistics and floating power expertise to meet the surging global demand for digital infrastructure through innovative floating data centers.

  • Get ready for a massive surge in AI-driven shopping this Amazon Prime Day 2025!

    It is expecting a 3,200% increase in traffic from generative AI sources compared to last year. Consumers are increasingly leveraging AI assistants like ChatGPT, Perplexity, and Amazon’s Rufus to find products, compare prices, and snag the best deals across various retailers. Adobe’s analysis further reveals that these AI-driven visitors are significantly more engaged on retail sites.

    Let’s have a look the impact on Amazon’s Infrastructure:

    • Scaling compute and storage: Amazon has historically prepared for Prime Day by significantly scaling its cloud infrastructure. For example, in 2022, Amazon increased Amazon EC2 compute instances by 12% and added 152 petabytes of storage to handle peak loads, processing trillions of requests and hundreds of billions of transactions daily. For 2025, with generative AI traffic expected to surge by 3,200% year-over-year, Amazon will have to further expand its AI-optimized compute resources, including GPU-powered instances and AI chips, to serve millions of real-time inference requests while maintaining low latency.
    • Advanced AI infrastructure: Amazon’s AI shopping assistant Rufus and other AI features rely on large language models (LLMs) that require highly efficient, scalable deployment to meet strict latency SLAs (e.g., 300 ms response times) during peak traffic. Amazon uses innovations like parallel decoding and specialized AI chips to improve inference speed and power efficiency, critical for managing the massive AI workload surges on Prime Day.
    • Automatic scaling and resilience: Services like Amazon Aurora automatically scale with traffic increases to keep checkout and other critical processes smooth and responsive. The infrastructure must handle not only raw traffic but also the complexity of AI-driven personalization and dynamic content generation without outages.

    What about the Impact on Customer Experience?

    • Personalized shopping at scale: AI-powered tools such as Rufus, AI-generated shopping guides, and interest-based recommendations aim to solve the long-standing challenge of deal discovery among millions of items. These AI assistants curate product selections, reducing choice overload and helping shoppers find relevant deals quickly.
    • Enhanced engagement and conversion: Adobe’s analysis shows AI-driven visitors stay 8% longer, view 12% more pages, and bounce 23% less than non-AI referrals, indicating deeper engagement and better-informed purchasing decisions. AI helps shoppers with product research, deal spotting, gift ideas, and personalized recommendations, improving satisfaction and increasing average order value.
    • Mobile and AI synergy: Mobile commerce accounts for over half of Prime Day sales, and AI-powered mobile shopping assistants are increasingly active, spotting real-time deal drops and enabling seamless, on-the-go purchasing. This integration of AI and mobile enhances convenience and responsiveness.
    • Sustainability and cost efficiency: AI optimizations also help reduce power consumption and operational costs, contributing to more sustainable infrastructure management during the intense Prime Day event.

    Amazon’s infrastructure and AI innovations are critical to delivering a seamless, personalized, and high-performance shopping experience during the largest and longest Prime Day ever, despite the unprecedented surge in generative AI traffic.

  • Expansion of Groq’s AI inference infrastructure into Europe

    Groq has partnered with Equinix to launch its first European data centre in Helsinki, Finland, marking a significant expansion of Groq’s AI inference infrastructure into Europe. This new facility aims to provide low-latency, scalable, and cost-efficient AI inference capacity closer to European users, enabling faster response times and stronger data governance.

    Let’s have a look Key details of this collaboration include:

    • Location: Equinix data centre in Helsinki, Finland, chosen for its sustainable energy policies, free cooling, and reliable power grid, making it ideal for hosting AI infrastructure.
    • Purpose: To support AI inference workloads—where trained machine learning models draw conclusions from new data—by delivering high capacity and efficiency at scale.
    • Customer benefits: Equinix Fabric customers can deploy inference workloads directly to GroqCloud, gaining access to inference capacity across the US and EMEA regions via public, private, or sovereign infrastructure options.
    • Strategic significance: This European footprint builds on Groq and Equinix’s existing collaboration in Dallas, Texas, strengthening their global AI infrastructure network and enabling customers to innovate faster in AI applications.
    • Statements: Jonathan Ross, Groq CEO, emphasized the growing demand for AI inference and the need for scalable, efficient infrastructure. Regina Donato Dahlström, Managing Director for the Nordics at Equinix, highlighted Finland’s advantages and the synergy between Groq’s technology and Equinix’s infrastructure.

    Groq’s investment in Equinix’s Helsinki data centre establishes a strategic European hub for AI inference, enhancing performance and accessibility for AI workloads across Europe and beyond.

  • CoreWeave announced its intention to acquire Core Scientific in an all-stock deal valued at approximately $9 billion

    CoreWeave and Core Scientific are closely linked players in the high-performance computing (HPC) and AI cloud infrastructure space, with a significant recent development being CoreWeave’s planned acquisition of Core Scientific.

    Let’s have a look Relationship and Business Focus:

    • Core Scientific is a leading provider of digital infrastructure, specializing in high-density colocation services and digital asset mining. It operates multiple purpose-built data centers across several U.S. states (Alabama, Georgia, Kentucky, North Carolina, North Dakota, Texas, Oklahoma) and is pivoting from Bitcoin mining toward supporting AI and HPC workloads through colocation services. It has contracts with CoreWeave to host GPUs and deliver AI-related compute capacity.

    • CoreWeave  is a cloud infrastructure provider focused on AI workloads, leveraging GPU-accelerated computing. It has been a major customer of Core Scientific, leasing significant data center capacity to run AI cloud services.

    Key Recent Development: Acquisition Deal

    • On July 7, 2025, CoreWeave announced its intention to acquire Core Scientific in an all-stock deal valued at approximately $9 billion. This deal is expected to close in Q4 2025, subject to regulatory and shareholder approval.

    • The acquisition will allow CoreWeave to gain direct control over Core Scientific’s data center infrastructure, including about 1 gigawatt of gross power capacity with an additional gigawatt planned for expansion. This will eliminate CoreWeave’s need to pay rent to Core Scientific on several data center sites, improving operational efficiency and saving an estimated $500 million annually.

    • The combined entity will have enhanced scale and capabilities to serve the growing AI and HPC market, with Core Scientific’s 12-year contracts with CoreWeave representing over $10 billion in recurring revenue.

    Strategic Shift and Financials

    • Core Scientific has been shifting its business model from Bitcoin mining, which has become less profitable, to high-density colocation services tailored for AI workloads. It plans to deliver 250 megawatts of billable capacity to CoreWeave by the end of 2025.

    • Despite a revenue decline in early 2025 (Q1 revenue $79.5 million, down from prior year), Core Scientific reported a net income of $580.7 million due to financial restructuring and strategic moves.

    • The partnership and upcoming merger reflect a broader industry trend of data center providers focusing on AI infrastructure, where demand for GPU-accelerated computing is rapidly growing.

    This acquisition positions CoreWeave to become a major integrated player in the AI infrastructure market by combining Core Scientific’s extensive data center assets with its own AI cloud capabilities.

  • The conflict between Google and the European Independent Publishers Alliance

    The conflict between Google and the European Independent Publishers Alliance centers on Google’s use of European journalistic content in its search services, particularly its AI-generated summaries called AI Overviews, and the fair remuneration of publishers under EU copyright laws.

    Let’s have a look Key points of the dispute:

    • Publishers’ complaint: The Independent Publishers Alliance accuses Google of abusing its dominant position in online search by misappropriating content for its AI Overviews, which appear prominently above search results. They argue this practice harms publishers by reducing traffic, readership, and revenue, effectively disadvantaging the original journalistic works.

    • Google’s unilateral actions: Publishers reject Google’s unilateral “experiment” in eight EU countries where Google removed press content from its services without consultation, affecting about 2 million Europeans. Google assessed the value of European press content using opaque criteria, which publishers say undermines their legitimate claims for fair payment.

    • Regulatory context: The European Union’s Digital Markets Act (DMA) and Copyright Directive seek to ensure fair remuneration for press publishers. The French Competition Authority intervened to prevent Google from delisting publishers during negotiations, safeguarding fair bargaining.

    • Google’s response and deals: Google has signed licensing agreements with over 2,600 European publications across 16 countries, including major deals in Germany and France, and is rolling out tools to facilitate contracts with smaller publishers. However, Google has also withdrawn from some deals, such as a multi-year agreement with Australian publishers, complicating relations.

    • Recent legal action: On July 4, 2025, the Independent Publishers Alliance filed an antitrust complaint with the European Commission, seeking interim measures to prevent further harm from Google’s AI Overviews, accusing Google of monopolistic abuse in search.

    This ongoing conflict highlights the tension between digital platforms’ AI-driven content aggregation and the rights and revenues of traditional news publishers in Europe, with regulatory bodies increasingly involved to ensure fair competition and remuneration.