Big Tech’s Thirst for AI is Unquenchable

Big Tech’s insatiable thirst for AI is becoming more prominent, with recent earnings calls standing as a testament to this trend.

Microsoft and Amazon have emerged as leaders, demonstrating significant growth and advancements in AI initiatives. The former reported a 16% increase in revenue , reaching $65.6 billion, with a notable emphasis on AI integration across its product suite. “Copilot is the UI for AI. Nearly 70% of the Fortune 500 now use Microsoft 365 Copilot,” CEO Satya Nadella highlighted.?

He said it has built an end-to-end system for AI business transformation with Microsoft 365 Copilot, Copilot Studio and agents, and now autonomous agents.

Nadella further pointed out Microsoft’s strategic approach to leverage commercial AI investments across LinkedIn, Windows, and gaming, adding that “Copilot+PCs ” herald a new era for hybrid AI at the PC edge, supporting developers and reinforcing its unified investment strategy.?

Amazon achieved a 11% revenue growth totaling $158.9 billion , while AWS saw a 19.1% increase, reaching a $110 billion annualised run rate, driven by strong adoption across cloud, AI and e-commerce.?

“Generative AI is a multibillion-dollar revenue business growing triple-digit year-over-year at a pace faster than AWS in its early days,” said Amazon chief Andy Jassy , underscoring the company’s strategic shift towards integrated AI services across multiple business segments.

Jassy also touched upon Bedrock , a foundational layer for deploying generative AI applications, which now offers “the broadest selection of leading foundation models”, allowing customers flexibility in model selection, while Amazon Q Developer , described as “the most capable generative AI-powered assistant for software development”, has helped save Amazon’s teams millions in productivity gains.?

“Customers are excited about the price performance that they believe they’re going to get in Trainium ,” said Jassy, underscoring the strategic role of Amazon’s custom silicon in optimising AI workloads cost-effectively and at scale.?

Meanwhile, Apple and Netflix have shown a rather modest performance. The Cupertino-based tech giant's revenue remained relatively flat at $94.9 billion . “We are just at the beginning of what we believe generative AI can do,” CEO Tim Cook noted, highlighting the long-term vision for AI integration across devices.??

Cook confirmed the integration of ChatGPT within Apple Intelligence . It enhances Siri’s conversational abilities and brings new AI functionalities to the Photos app and writing tools, allowing users to “create movies simply by typing a description”.??

Netflix reported a 15% increase in revenue to $9.8 billion but faces challenges in subscriber growth. Co-CEO Ted Sarandos remarked, “We are exploring new AI-driven content recommendations to enhance user engagement.”

Meta and Alphabet (Google) are in transitional phases. Meta’s revenue grew by 19% to $40.59 billion, thanks to strong product momentum and AI advancements. Meta chief Mark Zuckerberg said, “We estimate that there are now over 3.2 billion people using at least one of our apps each day, and we’re seeing rapid adoption of Meta AI and Llama.”?

Meta AI currently has over 500 million monthly active users. Notably, it has integrated new capabilities, including real-time multimodal interaction through Ray-Ban Meta smart glasses , where users can “see what you see, hear what you hear, and talk to you”.?

“Glasses are the ideal form factor for AI,” said Zuckerberg, highlighting how its glasses, equipped with Meta AI allows users to take photos, receive real-time translation and get contextual assistance.        

He also expressed excitement about Llama 4 , stating, “We’re training the Llama 4 models on a cluster that is bigger than 100,000 H100s.” He also hinted at releasing smaller versions of Llama 4 early next year .

Meta’s Reality Labs segment saw a 29% revenue growth to $270 million, though it reported a $4.4 billion operating loss, reflecting its continued heavy investment in AR/VR and infrastructure for long-term returns.

Alphabet reported a 15% revenue increase to $88.3 billion , with significant investments in AI research and development.?

Google and Alphabet CEO Sundar Pichai highlighted the company’s robust momentum in AI innovation across products and infrastructure, emphasising that AI search enhancements , like Circle to Search and Lens , now handle 20 billion visual searches monthly, while AI Overviews are available to over 1 billion users globally.?

Google Cloud also surged with 35% revenue growth to $11.4 billion, boasting 17% operating margins due to accelerated demand for generative AI and efficiency in TPU and GPU usage.        

Pichai lauded the Gemini’s performance , noting its broad use across all Google platforms, including YouTube , where ad revenue grew 12% as AI improves recommendations, enhancing engagement and monetisation across Shorts and in-stream video. Plus, Waymo , Alphabet’s autonomous vehicle division, reached a milestone with 1 million autonomous miles driven weekly.?

For more real-time updates and actionable insights, visit AIM .?

Best Firm for Data Scientists >>

Best Firm for Data Scientists is one of India's biggest workplace certification platforms in data science.?

In a first, IDFC First Bank and Aays have been certified as Best Firms for Data Scientists for the second consecutive year.?

To nominate your organisation for the certification, fill out the form here .

Top Stories of the Week >>?

AI is Killing Remote Work

AI advancements are reshaping remote-work dynamics, as Gumroad founder Sahil Lavingia said, “Software that once took days to ship can now be delivered in hours or minutes.” This is all thanks to AI tools like Claude 3.5 Sonnet, enabling workflows up to “10-20 times faster”. But, there are drawbacks. Read to find out.?

Anthropic is Now Big Tech’s Favourite Child—OpenAI, Not So Much

Anthropic is emerging as Big Tech’s hot favourite, with Claude outperforming in models and gaining integration across major platforms like GitHub Copilot and Amazon Q Developer , while Meta’s Llama also rivals ChatGPT in popularity . Meanwhile, OpenAI , initially a frontrunner, faces challenges with mixed feedback on its latest releases of o1. Has OpenAI lost its edge?

Dive into the full story for a deeper analysis.

AMD is Finally Catching up with NVIDIA Blackwell

AMD is rapidly closing the gap with NVIDIA , as its chief Lisa Su looks to introduce the upcoming MI325 and MI350 GPU accelerators to compete with NVIDIA's H200 and Blackwell series , respectively. Explore the full story for insights into AMD's AI ambitions.

People & Tech >>?

Setting Up Multimodal Pipelines and Vector Databases for Production Scale

As enterprises face rising data complexity, setting up multimodal pipelines and managing vector databases at production scale have become essential for real-time, data-driven applications.?

But, these are easier said than done. Challenges like latency, data consistency, and continuous optimisation persist, necessitating careful model tuning and infrastructure scaling. In this article, Chetan Dixit , client partner at Fractal, explains how multimodal pipelines unify different data types like video, audio, and text, storing vector embeddings in databases to provide customer support teams with a comprehensive interaction history. Learn more here.?

Tech Talks >>?

[Must Watch] Father of Indian Hardware: Dr Ajai Chowdhry | Spearheading the Quantum Mission in India

In the recent episode of Tech Talks, Ajai Chowdhry , often called the “Father of Indian Hardware”, shared his transformative vision for India’s journey from a service-led economy to a global product powerhouse.?

A co-founder of HCL, Padma Bhushan Dr Chowdhry recounted his pivotal role in establishing India’s IT hardware sector and highlighted India’s next big leap: self-reliance in electronics, semiconductors, and quantum technology.?

AI Nuggets >>?

• OpenAI finally launched SearchGPT in ChatGPT , offering ad-free, up-to-date web search with source-linked answers, challenging Google Search , Perplexity AI , and Bing for real-time information retrieval and user engagement.
• OpenAI also integrated chat history search in ChatGPT , powered by newly acquired Rockset , to enhance RAG with fast, contextual searches across user data—benefiting enterprise customers, reducing hallucinations and enabling accurate, context-rich results.
• OpenAI partnered with Broadcom and TSMC to develop its first in-house AI chip by 2026, diversifying its compute power with AMD’s MI300X chips alongside NVIDIA GPUs .?
• Google recently introduced ‘Learn About ’, an interactive learning tool positioned to rival Perplexity AI , offering deep-dives, quizzes, and exploration features to enhance knowledge extraction.?
• Microsoft silently introduced Copilot Vision , an AI feature in the Edge browser that, with user consent, can view and interact with the content of web pages, offering summaries, translations, and task assistance without disrupting workflow.

• GitHub Copilot introduced a multi-model feature allowing developers to select from top AI models like Anthropic’s Claude 3.5 , Google’s Gemini 1.5 , and OpenAI’s latest 01-mini and o1-preview , challenging rivals like Cursor and others .?
• Meta FAIR recently unveiled three advancements in robotics and tactile perception : Meta Sparsh , a versatile encoder for vision-based tactile sensing; Meta Digit 360 , an artificial fingertip with human-level touch precision; and Meta Digit Plexus , a unified platform for robotic sensor integration—aimed at empowering open source innovation across fields like medical research and manufacturing.
• Jio-backed TWO AI unveiled Sutra HiFi, a multilingual voice model enhancing natural voice interactions across 30+ languages.?
• Anthropic’s Claude 3.5 Sonnet recently introduced Visual PDF Analysis for documents under 100 pages , enabling extraction of context from charts, images, and graphs, while expanding document upload limits to 30 MB.?
• NVIDIA launched HOVER , a compact 1.5 million parameter neural network enabling humanoid robots to perform complex, natural movements, trained in NVIDIA’s Isaac simulation for efficient real-world applications.?
• NVIDIA recently powered xAI’s Colossus , the world’s largest AI supercomputer, with 100,000 Hopper GPUs and Spectrum-X Ethernet networking, enabling high-speed AI training, doubled to 200,000 GPUs.?
• AMD recently unveiled OLMo , its first open source 1 billion parameter language models, trained with 1.3 trillion tokens on AMD Instinct MI250 GPUs, offering customisable, scalable AI solutions and enhanced performance in reasoning and instruction-following.?
• Etched and Decart launched Oasis , an interactive AI world model generating open-world games frame-by-frame based on user inputs, with real-time capabilities powered by the upcoming Sohu ASIC chip . Try it out here.?
• Osmo showcased ‘scent teleportation ’, digitising and recreating the smell of plum through AI and gas chromatography, hinting at a future where digital memories may include scent.
• Disney recently established the Office of Technology Enablement, led by CTO Jamie Voris , to coordinate its use of AI and mixed reality across film, TV, and theme parks.