The AI industry isn't competing anymore - it's consolidating into what veteran tech journalist Steven Levy calls 'The Blob.' A massive interconnected web of partnerships, investments, and deals now binds Nvidia, Microsoft, Google, and OpenAI into a single money-and-compute machine that's reshaping how artificial intelligence gets built and deployed worldwide.
Ten years ago, Elon Musk helped create OpenAI specifically to prevent AI from falling under corporate control. Today, that same OpenAI is worth half a trillion dollars, and the entire AI industry has morphed into what WIRED's Steven Levy calls 'The Blob' - a single, interconnected machine of partnerships that makes traditional competition look quaint.
This week's blockbuster deal between Microsoft, Nvidia, and Anthropic perfectly illustrates how the Blob operates. Microsoft pumps at least $5 billion into Anthropic - a direct rival to its key partner OpenAI - while Anthropic commits to spending $30 billion on Microsoft's cloud services. Meanwhile, Nvidia invests in Anthropic, which agrees to build on Nvidia chips. The result? Anthropic's valuation rockets from $183 billion to $350 billion in just two months.
'We are increasingly going to be customers of each other,' Microsoft CEO Satya Nadella explained during the deal announcement, practically coining the Blob's unofficial slogan. The three CEOs didn't even bother meeting in person - these mega-deals have become so routine they're announced via video call.
But the Blob extends far beyond this trio. The Stargate initiative binds OpenAI, Oracle, Nvidia, SoftBank, and an Abu Dhabi investment firm with US government backing. Google maintains compute deals across multiple cloud providers while developing its own Gemini models. Even asking an AI to map these connections proved dizzying - it took GPT-5 over two minutes to process what Levy calls 'one giant circular money-and-compute machine.'
The driving force isn't conspiracy but necessity. 'I don't think any of us realized how real the scaling laws would end up being, and therefore how much compute these things would require,' one AI executive told Levy. Training and running large language models demands such massive infrastructure that even tech giants can't go it alone. 'We've all turned into infrastructure and construction companies,' the executive admitted.
