Will OpenAI's Eventual Collapse Squash U.S. Economic Growth?
OpenAI will collapse, that's my assumption. How large will the ripple effect be?
OpenAI is flaming out like a dying star. Whether its final stage is a white dwarf, neutron star or black hole remains to be seen. One thing is certain - OpenAI’s cash burn rate is unsustainable. Even if OpenAI has a wildly successful IPO and raises $100 billion, it can’t continue to burn $12 billon per quarter (September 2025 Q) and hope to exist in perpetuity. Sora alone is burning $12 million per day.
Further, OpenAI’s cash burn will significantly increase over the next 2 years or more. LLMs are reaching diminishing returns under the current training and inference paradigm. Frontier LLMs require more capital and more compute to make the same gains made a year ago. The low-hanging LLM fruit is gone.
OpenAI does not have the AI research chops to invent the next AI paradigm. AI breakthroughs are more likely to come from Google ( GOOGL 0.00%↑ ) DeepMind (speaking of DeepMind, check out the free documentary below which has gone viral). Heck, OpenAI has lost most of its senior talent over the past couple of years. If that’s not a leading indicator of pending doom, I’m not sure what is.
Will OpenAI’s flaming out have a huge ripple effect on the U.S. Economy? No. Most of OpenAI’s published projects and capital commitments have yet to materialize. Whatever datacenters Microsoft ( MSFT 0.00%↑ ) is in-process on will have capacity consumed by Azure customers running machine learning models and remote servers, the demand for which continues to grow.
I do expect the AI bubble to continue to deflate as we move through this December quarter earnings season.
Will someone acquire OpenAI? Who can afford to compete with Google? Microsoft? AI is not in MSFT’s DNA. Same goes for Oracle ( ORCL 0.00%↑ ), SAP ( SAP 0.00%↑ ) and every other Software company you can think of.