Financial market watchers recently have been speculating about the looming AI BUBBLE, and some market analysts are comparing it to the bust of the Dot-com bubble in the 2000s.

The AI Bubble is brewing, but the tera-watt media and market hype surrounding today's Generative AI doesn't want you to know that.

In fact, the AI market is in a "bubble" from both a financial and a technological standpoint, and no one in academia or in Big Tech has a fix for this. The bubble just keeps on expanding, and no one can stop it.

From a financial standpoint, the AI bubble is growing through market acquisitions of smaller AI startups by ravenous Big Tech companies. Although investors get some limited short-term returns when AI startups are bought out, for the past decade, greater long-term returns on AI investments have been bottled-up in the ever expanding financial bubble surrounding the stratospheric valuations of these Big Tech companies, who still haven't really figured out how to successfully commercialize AI to a mass market.

The tech companies Alphabet, Amazon, Meta Platforms and Microsoft have pledged a combined $325 billion in spending on AI for 2025 alone. And the scope of the returns on that investment isn't very clear.

What is also not too clear is how much return investors have gotten on the more than $200 billion invested in "AI" in the past decade.

One thing is clear, however. There cannot be too much commensurate returns as AI investment moves into the trillion dollar market very soon, because there is a technological reason why the industry is in a financial bubble.

Industry watchers have already noted an inflection point in the media hype: the perception that AI is moving toward a bubble and away from a gold rush has now found its way into the financial market sector.

Besides a few very narrow applications, like voice cloning and text-to-image generation, Generative AI has not developed into anything beyond a more sophisticated search engine. The AI bubble keeps growing, but beyond the media hype, AI systems are not really getting any more intelligent.

Because the fundamental technology behind the entirety of Big Tech's AI is basically on the wrong side of the thermodynamic physics equation, the Generative AI model itself has ballooned to such a mammoth size that AI developers at a wide range of companies are now increasingly frustrated with ever longer training sessions followed by ever longer periods of testing and development - pushing off returns on their investments farther into the future.

According to multiple sources, OpenAI's GPT 4 model has over 1.8 trillion parameters, which may seem to add up to more intelligence, but what it really just adds up to is a larger, more power-hungry model requiring increasingly longer periods of time to train for specific applications by AI developers.

Adding parameters has not made the model any more intelligent, since the number of parameters has nothing to do with how well the model generalizes, it only allows the model to cover a greater area of application.

Even as investors are shoveling hundreds of billions of dollars into anything related to "AI", the problem is that companies are claiming that they are using AI to "do things better", but there is no real evidence of this in the workplace. All of these claims are really nothing more than future promises build on fanciful hopes, based merely on faith in the tech industry. Of course, amplified by the media hype.

Surveys confirm that for many workers, AI tools like ChatGPT actually reduced their productivity, by increasing the volume of content needed to complete a given task, which must now be checked and corrected for the inevitable errors introduced by the tool.

And tech researchers have found that for workers who do become more productive, their critical thinking skills then declined, presumably because they were offloading the thinking onto the AI.

Then we get to the other colossal costs related to the AI bubble itself.

Microsoft, which had planned to be carbon neutral by 2030, reportedly had to revise that goal after its 2023 emissions were 30% higher than 2020. Google also has seen its carbon emissions increase as well and no longer claims to be carbon neutral, while pushing its net-zero emissions goal further into the future.

And this is the trend for many other companies with ties to the AI server infrastructure. The hyper-amplified media hype surrounding AI has accelerated the power consumption of AI data centers, providing justification for many companies to revise their environmental promises.

According to Goldman Sachs research, the major tech players are expected to funnel $1 trillion over the next five years into chip manufacturing for graphic processing chips and to build the data centers to power AI computing.

And all of those racks of high-end GPU chips and supercluster data centers require exorbitant amounts of electricity to run and huge amounts of freshwater to cool. Microsoft has plans in place to buy up all of the electricity from the revived Three Mile Island nuclear power plant, and Oracle recently announced plans to build a giga-watt data center for AI, powered by a trio of nuclear reactors. Meanwhile, OpenAI has pitched the White House on the necessity of 5 giga-watt data centers, which would be enough power for about two million homes.

Where will investors turn to when the AI bubble pops?

There is a new AI technology, just recently developed, that offers an alternative to the decades-old Generative AI that is driving the current AI bubble.

This AI is not only new, it is also fundamentally different than the equation-based algorithmic AI that all of Big Tech has been developing for  the past 25 years.

Nor does this new AI technology rely upon the ultra-high-end graphics processing chips currently propelling Nvidia to stratospheric valuations and forcing power grid operators to build nuclear reactors just to drive these new AI data centers.

This new form of AI is not based on arcane mathematical equations - it has been designed from the ground up on those elementary principles of natural intelligence that Nature evolved into us humans. So, what is its secret?

In essence, this new AI technology revolves around a novel form of organic memory, similar to the form of memory Nature evolved for our natural intelligence. Simply put, intelligence has less to do with the number of processors for data and far more to do with the storage and retrieval of that data.

Nature learned millions of years ago how to evolve natural intelligence by staying on the right side of the thermodynamic physics equation. The reason that all of Big Tech's AI is on the wrong side of that equation is this: The power dissipation of the human brain is on average about 18 watts, where your average desktop computer might consume from 400 to 900 watts.

But the power draw of just one of the high-end graphics processor chips needed for generative AI is 350 watts, and the new data centers being built for AI computing are estimated to require a hundred thousand or more of these GPU chips for their AI computation.

A giga-watt data center - that still cannot come anywhere close to exhibiting the same level of general intelligence as the 18 watt brain that Nature evolved for us humans.

There is an alternative solution...

After a decade of development, the AI technology from the First Frontier Project is now ready to help you with everything you need AI for.