Category Archives: Science

That is the question at times. So is it ‘What the eyes see’ or ‘What the iiii’s see’? Both are applicable and the setting is that is comes with a subjective view. That is often the case. But this is not about the particular. The article that passed my eyes was quite good and the subtext is “The AI hyperscalers will likely spend more than $1 trillion on data centers next year. Can they make enough money to sustain the infrastructure boom?”, the question reverberates as I have been asking that same question for some time. We all see the ‘investments’ that goes deep into the trillions, and no one seems to be worried about Return on Investment, a setting that is clearly asked in every boardroom in the world. And no one is willing to walk that question. So I grin from a distance and see all these people go “AI AI AI AI” and more of that. Then they all point at some newscast where President Trump states that “The golden age of AI” is upon us. But that simple statement ‘Golden age’ requires a return on investment. That is how it always goes as long as Ive lived and the term return on investment might be somewhat new, but the setting of that requirement was already old when a panting in 1639 was commissioned. It was then lost and found again and is now known as the Night Watch (a sketch by Rembrandt van Rijn) and the article starts rather strong with “When Jessica Wachter, a finance professor at the University of Pennsylvania’s Wharton School, wanted to assess AI’s impact on the economy over the next few years, she faced a long list of business and technical uncertainties. So she started with what she calls a “remarkable fact” that is not in question: A handful of so-called hyperscalers are investing huge amounts of money to build AI data centers. Instead of trying to predict how useful and widely deployed AI models will be, she simply asked how fast the hyperscalers’ earnings will need to grow to justify their spending through 2027, when—she and her collaborator estimate—expenditures will reach nearly $1.1 trillion. It’s a no-nonsense accounting approach to making sense of today’s historical AI buildout.” The source is (at https://www.technologyreview.com/2026/09/15/1144028/ai-infrastructure-boom-investment-bubble-risk/amp/) and it is called MIT Technology Review. We are also given “While the hyperscalers plan to spend trillions, total AI revenues will be around $150 billion to $200 billion this year, says Gary Gensler, who ran the SEC during the Biden administration and is now a professor at MIT’s Sloan School. “The challenge is that the spending does not have commensurate revenues yet. That’s a fact,” he says. “And then the question is, is that an investment that will be paid off in the future?”” From a distance (as I personally see it) it is a bundle of technology firms who have (on the books) trillions and they are using it to play ands of high risk poker and the world is allowing this, because if they lose it all they can write it off against their taxation, so the people basically pay for it all and I see it as a whole lot of nonsense because AI does not yet exist. I have written about this on several occasions. So, should it not be done? I cannot answer this, because Machine Learning and Deeper Learning (what I call Deeper Machine Learning, or DML) is a strong tool, and it could come with Large Language Models (LLM) if that setting is warranted and it was merely wrongly sold. It gave me the setting that court cases would reign over all this in 2026 and I was proven correctly. What we see now, is a clever use of predictive analytics on a much larger scale, but it is not AI, as such the Return on Investment needs to be strong. And as we see here (in this article) “At stake in that trillion-dollar question is the financial health of the giant AI companies and the overall US economy—the investments could soon balloon to around 3% of GDP. The answer could also determine the fate of the hugely expensive data centers themselves. No one really knows how profitable and useful these multibillion-dollar behemoths will be down the road. Though AI models have made dazzling progress over the last few years, it’s anyone’s guess how much compute capacity we will need. The technology could become more efficient and therefore less dependent on raw computational power. Or demand for AI products could slow, or customers could turn to cheaper models.” And whilst we think of the risk that ‘cheaper models’ give us, that setting might prove rather difficult, because when cost is pushed to make way for revenue, costing becomes a big thing and it really is a big thing. So when we are given “No one really knows how profitable and useful these multibillion-dollar behemoths will be down the road” the issue of return on investment will show its ugly head and that is the price part of this debate and no one is having it, because these boar members are all “We need AI and we need it now”, all whilst the return on investment is not proven and not shown anywhere. Don’t get me wrong. There are clear cases where a setting exists and options exist. I was shown the case of the issue of lost property and the stage was shown that from weeks, there is a setting where weeks could be turned into a setting where it could be done in under two hours. That is clear return in investment and for airports and bus terminals it could be a space saver. And from there we see interactive improvements. These are good ideas, even great ideas that when AI is finally here it will become powerhouses, but there is the setting that proper database work and LLM might do the trick. Clever programming that does not require AI. We got by just fine before this fake AI and whilst these snake oil vendors are so settled in ‘their’ AI, all whilst they are using the principles of predictive analytics and that is not AI.

So then we get to “The risks, both to investors and to the economy, have become even greater this year, as these AI companies have begun borrowing large amounts of money to build more and more data centers. Free cash flow—operating cash flow minus capital expenditures—is expected to soon dip into negative territory for the group. Even Alphabet, known for generating and hoarding huge amounts of cash, reports in the latest quarter that its impressive revenues of nearly $120 billion were devoured by AI infrastructure spending, leaving it with a free cash deficit of some $5.9 billion—its first shortfall since Google went public in 2004.” This is because I see another shortfall in the short term. Everyone is so driven towards data centers, whilst President Trump has driven the EU and other places away from the vendors of the United States and the term ‘data sovereignty’ is becoming more and more commonplace and whilst everyone is seeing these data centres and Stargate centres. It requires data and the EU is moving fast away from whatever Microsoft and Google are handing down towards their own centers not using software or hardware from the United States. The cloud act is now making that no longer an option. We get that from Politico, who gave us some time ago “Europe is actively trying to break its deep-rooted dependence on American big tech and cloud infrastructure. While major U.S. hyperscalers (like Amazon, Microsoft, and Google) still control roughly 70% of the European cloud market, public institutions and governments are shifting away from them”, as such 2027 might see a rather large turnabout and what happens to these data centres that are lacking data? You might think this is easy, but it all impacts the return on investment. 

So when we get to “Performance of the expensive GPU chips at the core of the data centers—such compute electronics represent some 60% of costs—is roughly doubling every two years or so. The pace of progress helps explain the increasing wizardry of the AI models, but it comes with a cost. Owners of AI data centers that come online this year and next will need to spend billions more on the next generation of chips by the end of the decade if they want to stay competitive. Without the investments, says Mihir Kshirsagar at Princeton’s Center for Information Technology Policy, the data centers risk becoming “hulks,” stranded assets “scattered all over the place.”

To put it bluntly: The AI companies need to start making a lot more money. And they need to do it fast. But juicing their earnings alone still won’t be enough to sustain their data-center investments for the long term.” And that is merely the beginning and I saw this roughly two years ago when I questioned the entire return on investment setting in all this and this article written by David Rotman does a good job, even more eloquent than I would have been. Although we basically say the same, this article does so a little better (and definitely more eloquent) then I could have written. So when you see the billions due next year, optionally over the next 2 years. Where is the return on investment? Because that is the question that is out there and as the IT field is changing and moving away from the United States, they too will see diminished revenue numbers. That much is certain, so where does this all stand? I am expecting a setting of actual AI to be a little over a decade away, it depends on certain factors and it also take in account a setting that I personally see (which might be wrong) but I feel that there is no AI, or as some call it true AI and I believe that requires a trinary data setting. As I see it it requires quantum computers (which exist) with shallow circuits (which is still in an early stage, as far as I know) and it requires a trinary coprocessor, which I call a Epsilon processor. These elements are required for an aI system, I set the system using a trinary coprocessor because that makes sense in a setting that is in part binary, that setting makes sense. We cannot merely push trinary systems through, there is will be a stage where they both need to exist. Later these systems are likely to be completely trinary, but that is merely my thoughts on the matter. And all this is still set towards the stages of return on investment. When you are considering this, how many billions are still required and who is willing to place this onto a systems that is unlikely to turn profit for a few years, optionally ver a decade. Who has that kind of money? There are a few, but are they willing to surrender that kind of money? The question might seem simple but the setting is not as straightforward as anyone thinks. And I saw this all along, so who gave you all the idea that the golden age of AI was here? Because that requires a massive revenue, or am I wrong?

Have a great day today, I’m now 90 minutes from Thursday and in Toronto it is now breakfast time. The idea to start the day with breakfast in Eggspectation on Bay Street is a little overwhelming for me at the moment. So you all have a good one and I will write to you in about 17 hours.