TECHNOLOGY
April 22, 2026 • 5 min read

AI infrastructure spending is accelerating across every sector. From a potential $60 billion AI coding acquisition to Google embedding Gemini into billions of browsers, today’s tech landscape is defined by who controls the full stack — from chips to consumer interfaces.

SpaceX Targets $60B Cursor Acquisition in AI Coding Push

BIG TECH

SpaceX has secured the right to acquire AI coding startup Cursor later this year in a deal valued at up to $60 billion, or alternatively pay $10 billion tied to ongoing joint work between the two companies under the “SpaceXAI” banner. Cursor, which has quickly emerged as a leading player in AI-assisted software development, is also reportedly raising $2 billion in fresh funding from investors including Andreessen Horowitz, Nvidia, and Thrive Capital.

The move signals that developer tools are becoming one of the most valuable battlegrounds in the AI economy. By aligning with Cursor, SpaceX is positioning itself to compete with OpenAI’s Codex and Anthropic’s Claude in the race to embed AI directly into software development workflows. The sheer scale of the potential acquisition — $60 billion — underscores how rapidly the market values AI-augmented coding capabilities.

Google Embeds Gemini AI Directly Into Chrome for 3.5 Billion Devices

ARTIFICIAL INTELLIGENCE

Alphabet has introduced two AI features that embed its Gemini assistant directly into the Chrome browser. The first, Skills for Gemini, lets users save custom AI prompts and reuse them across any webpage. The second updates Chrome’s AI Mode so merchant product pages open in a side panel alongside an active Gemini conversation session. Chrome commands roughly 65% of the global browser market across 3.5 billion active devices.

Gemini has grown to 750 million monthly active users — a 100-fold increase over two years. Google also launched Gemini as a native macOS desktop application this week. The distribution advantage is difficult to overstate: no competing AI assistant from Microsoft, Apple, or OpenAI can replicate this consumer surface area overnight. Separately, Adobe announced plans to introduce outcome-based AI pricing for its CX Enterprise suite, charging customers based on demonstrable business results rather than raw token consumption.

Google Signs Multi-Billion Dollar Cloud Deal With Thinking Machines Lab

STARTUPS

Mira Murati’s Thinking Machines Lab has signed a new multi-billion-dollar agreement to expand its use of Google Cloud’s AI infrastructure, including systems powered by Nvidia’s latest GB300 chips. The deal is valued in the single-digit billions and makes Thinking Machines among the first Google Cloud customers to access GB300-powered systems, which offer a 2X improvement in training and serving speed compared to prior-generation GPUs.

The deal reflects the fierce competition among cloud providers to lock in fast-growing frontier AI labs. Anthropic separately signed agreements with both Google and Amazon this month, securing multiple gigawatts of TPU and general compute capacity. Thinking Machines, founded by the former OpenAI CTO in February 2025 and valued at $12 billion, launched its first product Tinker in October — a tool that automates the creation of custom frontier AI models using reinforcement learning.

Google’s Custom AI Chips Begin to Pressure Nvidia’s Market Dominance

HARDWARE

Investors are paying closer attention to Google’s next-generation Tensor Processing Units as a real competitive threat to Nvidia, according to Barron’s. While analysts still argue Nvidia retains a strong lead owing to its hardware roadmap and CUDA software ecosystem, the market reaction shows the AI chip battle is no longer purely hypothetical. Hyperscalers are building custom silicon to reduce dependence on a single dominant supplier.

If Google proves it can deploy homegrown chips efficiently at scale, it strengthens the case for a more fragmented AI hardware market where cloud giants capture more value internally. Nvidia rival Cerebras has also disclosed a US IPO filing, further evidence that the competitive landscape for AI silicon is broadening significantly. The most serious threat to Nvidia may ultimately come from its own largest customers turning into chip competitors.

Meta to Track Employee Keystrokes and Screen Activity for AI Training

PRIVACY

Meta is preparing to collect detailed workplace behaviour data from US-based employees, including mouse movements, clicks, and keystrokes, using new software installed on work devices. Reuters reported that the programme is tied to Meta’s effort to build AI agents capable of performing real-world tasks, turning routine digital activity into training material for future models.

The move signals a shift in AI training data from static text to live human behaviour, raising significant labour, privacy, and governance questions. The next phase of enterprise AI is no longer about scraping the web or licensing content — it is about capturing workflows, judgement, and human-computer interaction at scale. Meta is effectively treating ordinary workplace activity as a strategic AI asset, a precedent that could reshape how every large company approaches employee data.

ZY Media Productions

IT • Music • Technology