AI – Has the AI Bubble Reached its Frothy Top? Oracle suggests so.

Blain’s Morning Porridge 11th Sept 2025: AI – Has the AI Bubble Reached its Frothy Top? Oracle suggests so.

 “Sam Altman, the CEO of Open AI, like all the great tech barons of our era, is an adman.”

The AI sector looks likely to be the biggest bubble in market history – until the next one. When it pops – and it surely will – the effects will resonate round markets. Like all good Snake Oil scams, AI has promised everything and claims divinity. Too many folk have embraced the hype and hopes of easy money – but that has always been the way of markets.

My promise to readers is the Morning Porridge will never ever be written by AI. It will remain originally erratic thinking triggered by the events around me every time.

This morning, it’s time for a big question that is going to upset a lot of heavily invested people. Is the AI bubble bursting?

A few weeks ago I was ambushed on air by a question about Oracle – how big a beneficiary from AI might the venerable database company become? I was too proud to admit I knew little, but I’d read about the firms efforts to align its cloud business with the proliferation of AI data centres, and that there is a shortage of everything in the AI sector, so I hedged with a comment about older firms reinventing themselves to find a niche in the new AI-Everything market, but that the opportunity was time-limited. I rambled on about corporate evolution and how quickly undersupply can become flooded. Blah-bitty blah blah…

Yesterday’s Oracle stock rallied 43% – propelling Larry Ellison to World’s richest man. I’m thinking the $300 bln cloud deal it’s just done with Open AI (a company that doesn’t expect to make any significant profits till the end of this decade!) has got to signal the AI market’s frothy top.

Yet, this morning I’ve got emails asking if it’s the next Nvidia and how much they should buy. Please. Stop. Look. Listen.

Something wicked this way comes. The signals are mounting. My threat board flashed the day OpenAI’s Sam Altman appeared for a signing with Donald Trump and Masayoshi Son of Softbank in The White House to announce the $500 bln future of AI. The question of who is going to provide the cash Altman needs into a new firm, Stargate…. was fudged. Since then we’ve heard little. My brain has been screaming WeWork, WeWork, WeWork ever since….

 On the face of it, Oracle has staged a stunning success winning a number of new cloud computing contracts, but OpenAI’s is the largest by miles. It leaves Oracle somewhat over-exposed to a single customer with multiple issues to resolve. Not just OpenAI’s tieup with Softbank for finance, (Danger, Danger Will Robinson, Danger), but competition in the AI space, its efforts to spin itself into a for-profits company, and CAltman firing off new business ideas like Elon Musk after a hefty engagement with uppers. I wonder where his AI Phone idea, with former iPhone designer Jonny Ives, has gone?

The AI revolution has spawned the greatest investment boom in human history as investors scrabble to own the future. It is seen across markets as the big tech firms offer packages worth hundreds of millions to AI “experts”, or pay billions for AI start-ups with the merest kernel of an untested conceptual application. How many trillions have the big Tech firms invested into AI – its almost a competition to spend the most.

There are estimates that nearly half the $500 billion per annum invested in Private Equity or Venture Capital in the last 2 years has flowed into ventures that claim to be driven and based on Artificial Intelligence. In the less than transparent private capital markets, there is a scramble underway to fund the accoutrements of AI – data centres, and even massive bulk chip purchases by firms that are little more than start-ups. (Private credit bankers are secured on the underlying chips. Who will buy them when push comes to shove, and the market is flooded? China? Interesting…)

If I were Ellison sitting in Oracle HQ, I’d be wondering how best to hedge against chronic over-capacity in computing and data-centres if/when it pops.

All bubbles burst when the money runs out. Where will the money come from to sustain this one? Check out the Dot.com bubble of 2000 for what happens when the market is all over an idea, and suddenly realises it’s never going to make the profits founders have promised their investors!

When this one pops there is going to be loads and loads of deep remorse. What do all these financial genius and investment superstars think they are buying?

There is increasing scepticism about what AI actually is, but even so I don’t think markets are ready for the big reveal – AI is not intelligence… its marketing hype.

My opening line quote this morning about Sam Altman as an “Adman” is from a recent book on AI – The AI Con: How to fight Big Tech’s Hype.. It perceives AI as vested interest: “A few major well-placed players poised to accumulate significant wealth by extracting value from other people’s creative work, personal data, or labour, and replace quality services with artificial facsimiles.”

Meanwhile, for every person still trying to pump AI investments, there is someone serious on an investment desk starting to ask meaningful questions like – yes, but what is AI, what is it actually achieving, and where are the profits? It’s terribly familiar to veterans of the Dot.Com boom and bust. Its reaching that moment where smart money hopes not the be the greatest fool – the last buyer before the pop.

The clearest sign of an imminent bust is the sheer number of emails, linked-in adverts, and social media interruptions I receive each day offering to integrate AI into my business, my life, and my very being. I did a serious AI foundation course recently – it was laughable. There is increasing desperation of the AI complex to demonstrate its utility. After 40 years in financial markets I barely understand the complexity driving the interaction of sentiment, mood and function called the market that sets the price of financial assets, so of course I’m going to pay some spotty 20-year AI marketer thousands of pounds to let algorithms with go-faster stripes make my job easier for me, by giving AI firms all my data and information.

Not.

But AI has spread its tubers – listen to what your colleagues are actually saying about AI, and how they repeat all the key talking points. We have all heard so much about how it is going to change our lives. From politicians boasting of how they will harness the power of AI into economic growth, from businesses desirous of cutting costs threatening job losses, or frustrated office workers looking to escape drudgery, the promise that AI will make things better is enormous, but what will it actually deliver? Everyone talks ever so knowledgeably about upside expectation, and use cases, but most barely understand or comprehend what AI actually is.

The reality is AI is not intelligence.

But it is artificial. The AI Con describes ChatGPT as a Synthetic Text Extruding Machine. What it represents in the widest sense is better processing – a process of constant refinement since Alan Turing invented the computer. So yes, it is creating new products – like photo apps that can make fake porn, or little steps forwards towards self-driving cars. But the reality is businesses are losing customers because they’ve sacked call-centre customer support staff and replaced them with useless AI chatbots. How monetisable is it – really?

The hype promises transformational business opportunities and gains. Classic overpromise and underdelivery. In its current form AI is not infallible and it’s a bit stupid. One day it may be perfected, but by then it will just be a business choice, not a FOMO necessity.

The real genius of the brains behind AI was the marketing blitz around the launch of ChatGPT. The narrative noise was incredibly attractive, creating the illusion of a sudden paradigm leap forward in the capability of computers to process data – the message was buy in or lose out. FOMO squared. The trick was not calling it data processing – but elevating the mundane to the sublime by calling it Artificial Intelligence. (Play the ethereal rising chord sequences of a revelatory experience.)

AI has been presented as a revolution in computing and understanding. But, in truth it might have been little more than a brilliant marketing wheeze.

I parroted much of what I read. How there were less than 100 computer scientists who truly understood the tau of AI, and how valuable they were. I bought into the them it was transformational. But, in reality, the emergence of large language models that can parse the enormous content of the internet into coherent, logical answers written in concise English is not computational magic – it’s the application of computing power, and part of the long-term evolution of data technologies which are predictive machines.

There is no thinking in AI. Artificial Intelligence is just a continuation of a long process of finding better, faster and more user-friendly ways of scraping information from the internet, or whatever other data set you have to hand. The trick was selling it – so why not tell firms that it will save them money by allowing them to sack staff doing repetitive tasks, or analysing data. Yep – that works.

Now. I might be wrong. Maybe there is a spark of consciousness lurking at the bottom of a hard drive, or an advanced chip – but I seriously doubt it. That is what the AI Schills would like us to believe – it would further magnify the transformational hype. Scaring us with p(doom) scenarios (that AI will kill us all) is part of the hype. What I reckon is increasing processing power and better algorithms have made the unimaginable scale of data mountains manageable. It is not intelligence, it’s mathematics.

And it can make industry and business more efficient and effective without destabilising society through mass middle-class professional layoffs…

But people will always fall for Snake Oil.

The brilliant marketing of Artificial Intelligence has created perhaps the biggest investment bubble of all time. According to ChatGPT (I said I would not let it write the Porridge, but I will use it), the value of the top 10 AI firms has doubled to $16 trillion since Nov 22. Thats 120% upside. Nvidia is the obvious winner – up 750% in the same time frame as everyone seeks the chips apparently necessary for AI success. Is Oracle the next wonder stock – sure, if you can suspend the rules of credit and investment! (Why hype often does.)

All that investment into AI has squeezed out investment into other sectors – what happens if the whole AI sector goes boom?

I rather think it will…

Out of time and back to the day job…

Bill Blain

CEO – Windshift Capital

Author – The Morning Porridge

Partner – Shard Capital