Capital, Compute, and Cold Wars

What happened
Reuters reported that OpenAI is targeting roughly $600 billion in total compute spend through 2030 as it lays the groundwork for a potential public offering. The ChatGPT developer generated $13 billion in revenue in 2025 while spending $8 billion. The article noted that Nvidia is close to finalising a $30 billion investment into OpenAI as part of a >$100 billion fundraising round, valuing the startup at about $830 billion. Microsoft‑backed OpenAI expects more than $280 billion in revenue by 2030, split roughly equally between consumer and enterprise products.

Why it matters
The numbers underscore the staggering capital intensity of AI. If completed, the round would be one of the largest private capital raises ever, cementing OpenAI as a near‑trillion‑dollar entity and deepening ties between model makers and chip providers. Nvidia’s investment signals that hardware companies may increasingly take direct stakes in leading AI developers to secure demand for their chips.

What’s next
 Expect other investors to follow, pushing valuations higher and drawing regulatory scrutiny. The compute race could accelerate consolidation between cloud providers, semiconductor makers and AI labs, while the sheer scale of the planned spend will likely influence global data‑centre construction and energy policy.

What happened
During a staged ‘unity’ photo at the AI Impact Summit, Prime Minister Narendra Modi asked executives to raise their fists together. OpenAI’s Sam Altman and Anthropic CEO Dario Amodei—the leaders of two fiercely competing AI labs—kept their fists apart while their peers joined hands. The awkward moment went viral, epitomising what social media users dubbed an ‘AI cold war’ between the companies. It occurred as the summit sought to project harmony among global tech leaders.

Why it matters
The incident captured the escalating rivalry between OpenAI and Anthropic, whose founders split over disagreements about commercialisation and safety. While the summit promoted cooperation, the reluctance to link arms suggested that commercial competition remains intense—even on a world stage.

What’s next
Expect continued jockeying between the two firms as they vie for investors, talent and political favour. Their uneasy coexistence at high‑profile events may pressure organisers to navigate corporate rivalries more carefully in future.

What happened
Alphabet’s Google and Southeast Asian tech conglomerate Sea announced a strategic partnership to develop AI tools for Sea’s e‑commerce and gaming businesses. The companies will work on an “AI agentic shopping prototype” for Shopee and deploy AI solutions to boost productivity in game development. The move follows Alibaba’s release of a new model designed for the “agentic AI era”,

Why it matters
The partnership underscores how leading tech firms are racing to monetise AI beyond chatbots. By building agents that handle tasks across apps, Google and Sea hope to dominate Southeast Asia’s retail and gaming markets—areas where Shopee already holds a 52% market share.

What’s next
Expect pilot launches of agentic shopping assistants and increased competition as Alibaba, Amazon and other players introduce their own agents. The success of these prototypes will test whether consumers trust AI to manage purchases and navigate multiple platforms on their behalf.

What happened
Beyond the pageantry and problems, the India AI Impact Summit triggered a flood of investment pledges. Reliance Industries and Jio committed $109.8 billion to build AI and data infrastructure over seven years, while the Adani Group vowed to spend $100 billion on renewable‑powered AI data centres. Microsoft said it will invest $50 billion by 2030 to expand AI access across the Global South. Yotta Data Services announced a >$2 billion AI compute hub using Nvidia’s latest chips, and Larsen & Toubro teamed with Nvidia to build India’s largest AI factory.

Why it matters
The sheer scale of these commitments positions India as a major player in the global race for AI infrastructure. By pledging sums rivaling national GDPs, Indian conglomerates and global tech firms signal that AI data centres, chip factories and cloud platforms will be essential economic assets—analogous to oil refineries in previous eras.

What’s next
Delivering on these promises will require navigating land acquisition, energy supply and regulatory hurdles. If realised, the projects could anchor a South Asian AI ecosystem that draws developers and capital from around the world.

What happened
ByteDance dropped Seedance 2.0, a next-gen video generator that lets anyone create viral clips starring Hollywood A-listers—no studio sign-off required. Within hours, Netflix and other studios fired off cease-and-desist letters, accusing ByteDance of trampling on IP and celebrity likeness rights. ByteDance admitted the model launched with “minimal guardrails” and promised fixes, but the genie is out of the bottle.

Why it matters
This is the copyright clash everyone saw coming. Generative AI is now powerful enough to blur the line between fan fiction and blockbuster, threatening traditional content licensing and creator revenue. The legal pushback signals a new era of IP battles, with studios scrambling to defend their turf as AI democratizes content creation.

What’s next
Expect a wave of lawsuits, tighter AI content moderation, and maybe even new laws. Meanwhile, creators and platforms will race to balance innovation with compliance—while the internet drowns in AI-generated celebrity mashups.

What happened
Agentic AI—autonomous agents that plan, remember, and use tools—hit a tipping point. Anthropic’s Model Context Protocol (MCP) became the “USB-C for AI,” adopted by OpenAI, Microsoft, and Google, and donated to the Linux Foundation. Gartner now predicts 40% of enterprise apps will embed AI agents by mid-2026, up from less than 5% last year.

Why it matters
Agentic AI is moving from research labs to real business workflows, automating everything from customer support to scientific discovery. Standard protocols like MCP mean agents can finally talk to each other—and to your apps—without friction.

What’s next
Expect a flood of agentic tools in enterprise software, new security and governance challenges, and a race to build the “operating system” for autonomous workflows.

AI is no longer a software story — it’s a capital-intensive infrastructure race spanning chips, data centers, geopolitics, and enterprise integration. The winners won’t just build better models — they’ll control the compute, partnerships, and standards that power them.