AI is the biggest bubble in the history of tech. I think the peak will be in mid to late 2027 with a 90% collapse in valuation in the next 12 months after that.
There is already a potent trigger for an AI bust. AI research has shown for some years that scaling LLM training has reached a limit. ChatGPT5 is hardly any better than ChatGPT4.0, and all the AI companies like OpenAI are incredibly unprofitable. People are starting to realise that the reality of what AI can do is far short of the promises that have been made. And yet the amount of money that is going into AI is genuinely extraordinary. Far more than the Dot Com boom. By late 2007, investors will have lost patience and companies will have run out of money. And that's when the bust will start, if not earlier.
AI is the biggest bubble in the history of tech. I think the peak will be in mid to late 2027 with a 90% collapse in valuation in the next 12 months after that.
But there must be some solid trigger point for the bust? In 2000 it was inflated companies IPOs, 2008 it was CDS fraud.
There is already a potent trigger for an AI bust. AI research has shown for some years that scaling LLM training has reached a limit. ChatGPT5 is hardly any better than ChatGPT4.0, and all the AI companies like OpenAI are incredibly unprofitable. People are starting to realise that the reality of what AI can do is far short of the promises that have been made. And yet the amount of money that is going into AI is genuinely extraordinary. Far more than the Dot Com boom. By late 2007, investors will have lost patience and companies will have run out of money. And that's when the bust will start, if not earlier.