A Daily Chronicle of AI Innovations on March 20th 2025
Nvidia announced substantial investments in US manufacturing and acquired Gretel to enhance its AI cloud services. Simultaneously, SoftBank acquired chip designer Ampere, strengthening its AI infrastructure portfolio. Apple restructured its AI leadership with the goal of improving Siri. Furthermore, AI capabilities were reported to be accelerating beyond the pace of Moore's Law, while Hollywood creatives voiced strong opposition to proposed AI copyright exemptions. Finally, Nvidia open-sourced reasoning models and launched personal AI supercomputers, and Hugging Face released an AI-powered image description app.
💥 Nvidia to Invest Hundreds of Billions in US Manufacturing
Nvidia CEO Jensen Huang announced plans to invest several hundred billion dollars in U.S. semiconductor and electronics manufacturing over the next four years. This strategic move aims to shift Nvidia’s supply chain from Asia to the United States, aligning with recent U.S. policies promoting domestic production. The investment underscores Nvidia’s commitment to bolstering its manufacturing capabilities within the U.S.
The company aims to manufacture its latest systems in the US by collaborating with suppliers like TSMC and Foxconn, aligning with other tech firms' US investment strategies.
Additionally, Nvidia has acquired synthetic data firm Gretel for over $320 million to enhance its cloud-based AI services, addressing data scarcity and privacy issues.
What this means: Nvidia’s substantial investment is poised to strengthen the U.S. semiconductor industry, enhance supply chain resilience, and support the growing demand for AI technologies.
💰 SoftBank Acquires Chip Designer Ampere for $6.5 Billion
SoftBank Group has agreed to acquire Ampere Computing, a U.S.-based chip design startup, for $6.5 billion in cash. Ampere specializes in high-performance, energy-efficient processors for cloud computing and AI workloads. This acquisition aligns with SoftBank’s strategy to expand its footprint in the AI infrastructure sector.
Oracle and Carlyle, the primary shareholders in Ampere, have agreed to relinquish their stakes, allowing Ampere to become an independent subsidiary of SoftBank.
Ampere, which focuses on creating Arm-based processors for data centers, will maintain its operations in Santa Clara, California, and employ about 1,000 semiconductor engineers.
What this means: SoftBank’s acquisition of Ampere enhances its position in the AI and semiconductor markets, potentially accelerating advancements in AI computing capabilities.
👀 Apple Reshuffles AI Leadership to Improve Siri
Apple has undertaken a significant restructuring of its AI leadership, aiming to enhance Siri’s capabilities and overall AI strategy. The reshuffle involves appointing new leaders with expertise in machine learning and AI to drive innovation and address competitive pressures in the voice assistant market.
Mike Rockwell, who previously led the Vision Products Group, will now head the Siri team, while John Giannandrea will remain to focus on research and other AI technologies.
Paul Meade will succeed Rockwell in leading the Apple Vision Pro team, having previously been in charge of hardware engineering for the same product line.
What this means: Apple’s leadership changes indicate a renewed focus on advancing Siri’s functionality, potentially leading to more natural and intelligent user interactions in future updates.
⚡ AI Capabilities Surpassing Moore’s Law
AI innovation is accelerating at a pace far beyond Moore’s Law. While Moore’s Law predicts that transistor density doubles every two years, AI model performance has been doubling roughly every six months. This phenomenon, sometimes called “Hyper Moore’s Law,” is driven not only by hardware improvements but also by advances in algorithm efficiency and data scale.
What this means: The exponential acceleration of AI could soon surpass traditional computing limitations, leading to rapid breakthroughs in AGI, reasoning, and real-time cognition. This trend will pressure infrastructure, education, and ethics frameworks to keep up with capabilities growing at super-Moore’s scale.
🎬 Hollywood Pushes Back Against AI Copyright Proposals
Hollywood actors, screenwriters, and musicians are uniting to oppose AI companies like OpenAI and Google, who are lobbying for copyright exemptions that would allow AI models to train on protected creative works without permission. Over 400 prominent creatives, including Aubrey Plaza and Paul McCartney, are demanding that the U.S. government prevent what they call “AI theft of creative labor.”
What this means: The copyright vs. fair-use debate is reaching a tipping point. If AI companies win exemptions, it could destabilize the business model of creative industries. On the flip side, strict restrictions could slow AI progress, particularly in generative content and cultural modeling.
🧠 Nvidia Open-Sources Reasoning Models to Advance AI Collaboration
At GTC 2025, Nvidia announced it is open-sourcing its advanced reasoning models, including new architectures focused on multimodal understanding, symbolic reasoning, and agentic task decomposition. These models aim to enable developers and researchers to build more capable, human-aligned AI agents.
What this means: By releasing foundational reasoning models to the public, Nvidia is democratizing access to next-generation AI capabilities. This could accelerate AGI development, foster academic collaboration, and open doors for startups to build on top-tier AI infrastructure.
🖥️ NVIDIA Announces DGX Spark and DGX Station: Personal AI Supercomputers
At GTC 2025, NVIDIA unveiled DGX Spark and DGX Station, a new line of personal AI computing systems designed for individuals, researchers, and developers who require on-premises AI power. These machines are equipped with the latest Blackwell architecture, offering workstation-sized compute power previously limited to data centers.
What this means: AI development is becoming personalized and decentralized. With DGX Spark, AI researchers, startups, and creators can now build and fine-tune models locally, potentially reducing reliance on cloud GPUs and unlocking a new era of personal-scale AI innovation.
📱 Hugging Face iOS App Uses AI to Describe What You See
Hugging Face has launched an iOS app that uses AI-powered image recognition to describe what users are looking at in real time. The app leverages open-source vision-language models to generate detailed and contextual descriptions, assisting both visually impaired users and general consumers interested in image-based discovery.
What this means: This move puts multimodal AI tools directly into consumers’ hands, with practical accessibility applications and the potential to become a daily tool for education, travel, and discovery. It also highlights Hugging Face’s role in making open-source AI usable at scale.
🧠 Optimizing Generative AI Using Language Model Feedback
New research published in Nature reveals a technique to optimize generative AI outputs by backpropagating feedback from language models themselves. Instead of relying solely on human annotation, the system uses AI-generated evaluations of output quality to refine the generator’s performance.
What this means: This method could reduce dependence on costly human feedback in reinforcement learning (RLHF) while enhancing model quality. It opens the door to self-improving AI systems that continuously refine outputs based on their own assessments, pushing us closer to autonomously evolving AI agents.
🌮 AI Will Soon Take Your Order at Taco Bell and Pizza Hut
Yum! Brands is rolling out AI-driven voice assistants at Taco Bell and Pizza Hut drive-thrus across the U.S. These systems, powered by conversational AI and large language models, can take orders, upsell, and handle payment interactions without human involvement.
What this means: Fast food is entering the age of AI automation, aiming to improve order accuracy, reduce labor costs, and speed up service. However, it also raises concerns over job displacement and the reliability of AI under real-world pressure.
What Else Happened in AI on March 20th 2025?
Google AI and UC Berkeley researchers proposed "inference-time search" as a new AI scaling method, producing several answers in parallel and selecting the best option.
LG released EXAONE Deep, a reasoning AI that achieves comparable performance to models like DeepSeek V1 in math, science, and coding with just 32B parameters.
Muse released Muse S Athena, a headband wearable combining EEG and sensors to measure both brain activity and oxygen levels for AI-powered cognitive fitness training.
Nvidia and xAI are joining Microsoft, BlackRock, and MGX in the AI Infrastructure Partnership, aiming to raise $30B initially and potentially $100B for AI data centers.
xAI debuted its first image generation API featuring the ‘grok-2-image-1212’ model, allowing developers to create multiple JPG images per request at $0.07 each.
Microsoft is partnering with neuroscience AI startup Inait to develop brain-inspired AI that learns from real-world experiences rather than data patterns.
🚀 From Our Partner:
Djamgatech's Certification Master app is an AI-powered tool designed to help individuals prepare for and pass over 30 professional certifications across various industries like cloud computing, cybersecurity, finance, and project management. The app offers interactive quizzes, AI-driven concept maps, and expert explanations to facilitate learning and identify areas needing improvement. By focusing on comprehensive coverage and adapting to the user's learning pace, Djamgatech aims to enhance understanding, boost exam confidence, and ultimately improve career prospects and earning potential for its users. The platform covers a wide array of specific certifications, providing targeted content and practice for each, accessible through both a mobile app and a web-based platform.
📥 Get Djamgatech (iOs) at Apple App Store: https://apps.apple.com/ca/app/djamgatech-cert-master-ai/id1560083470
🚀Advertise on AI Unraveled: Reach Thousands of AI Enthusiasts Daily!
AI Unraveled is your go-to podcast for the latest AI news, trends, and insights, with 500+ daily downloads and a rapidly growing audience of tech leaders, AI professionals, and enthusiasts. If you have a product, service, or brand that aligns with the future of AI, this is your chance to get in front of a highly engaged and knowledgeable audience. Secure your ad spot today and let us feature your offering in an episode!
🎙️ Book your ad spot now: https://buy.stripe.com/fZe3co9ll1VwfbabIO
🙏 Support the AI Unraveled Podcast and Channel:
Consider buying me a coffee to say thank you for the free tech content on my YouTube channel (@enoumen) and the AI Unraveled podcast. https://buy.stripe.com/3csaEQ1ST9nYgfe4gk
Share this post