The race in artificial intelligence is intensifying. Elon Musk's SpaceX, which develops AI models alongside space rockets, announced last week it is seeking a $1.77 trillion valuation on the US stock market. Meanwhile, Anthropic, the startup behind the Claude chatbot, has filed for an initial public offering, and OpenAI, the developer of ChatGPT, is expected to follow.
AI Drives Stock Market Surge
The S&P 500 has risen nearly 80% over the past five years, fueled by big tech stocks involved in AI, known as the "magnificent seven": Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla. According to Jim Bianco of Bianco Research, 41 AI-related stocks now account for nearly half of the S&P 500's market value, an unprecedented concentration. Analyst Neil Wilson of Saxo UK warns of a potential repeat of the dotcom bubble, calling the market "incredibly dangerous."
Expenditure on AI Infrastructure Skyrockets
Spending on AI infrastructure, from datacentres to chips, is projected to grow from $765 billion this year to $1.6 trillion by 2031, according to Goldman Sachs. However, the bank acknowledges risks: delays in datacentre construction could undermine demand assumptions. Despite this, the scale of investment reflects enormous global financial commitment and expectations for returns.
Adoption Accelerates Among Firms and Consumers
McKinsey reports that nearly 80% of companies now use AI, up from 33% in 2023. ChatGPT has reached 1 billion monthly active users, a record for any app. The challenge for AI developers is monetizing this user base by demonstrating that AI improves outcomes and reduces costs, enabling entire workflows to be automated.
Claude Challenges ChatGPT
Anthropic's Claude Code tool has gained traction, especially among developers, and represents a shift toward autonomous AI agents. Data from Kentik shows Claude's user traffic growing faster than ChatGPT's and Google's Gemini between January and April. At this rate, Claude could overtake ChatGPT by summer, making Anthropic's IPO path easier.
Rising Costs of AI Usage
AI costs are measured in tokens, with OpenAI charging $5 per million input tokens and $30 per million output tokens for GPT-5.5. These costs are rising, even as companies encourage employees to "tokenmaxx." AI companies still aren't charging enough to cover costs, undermining the promise of productivity gains. Liam Betsworth of Pendra notes that costs are "getting completely out of control," citing a company that spent $500 million in a month on Claude Code licenses.
Datacentre Construction Faces Challenges
Datacentre capacity must keep pace with AI demand to avoid a compute crunch. Bloomberg estimates 23 GW of capacity under construction globally in 2025, while JLL predicts 100 GW will be added between 2026 and 2030, equivalent to 1,200 datacentres. However, questions remain about funding and energy supply. Cecilia Rikap of University College London asks whether governments have calculated the feasibility and environmental impact of such expansion.
AI Capabilities Improve Rapidly
According to METR, AI models double in capability every four months, based on coding task completion times. However, this has not yet translated into significant job displacement. Bouke Klein Teeselink of King's College London notes that bottlenecks remain, but predicts huge changes ahead as many tasks could be automated.
Datacentres Prop Up US GDP
Despite government layoffs and industry cuts, US GDP grew 2.1% in 2025 and 1.6% in Q1 2026. A Harvard economist calculates that investment in information processing equipment and software accounted for 92% of US GDP growth in the first half of 2025, highlighting the disproportionate role of datacentres and AI in the economy. Any reduction in this spending could have significant economic and political consequences.
