Words Matters

The week ended on July 17, 2026, interesting news items to look at for the past fortnight are:

Words Matters: There is a saying by Confucius “一言兴邦,一言喪邦” which can be translated as, “A single remark can revitalize a state; a single remark can ruin a state.” (Fig. 1).

"A wise leader speaks with humility and welcomes honest counsel—then the nation prospers. An arrogant leader who shuts out dissent when wrong leads the nation toward ruin."

Trump and his team, of course, do not necessarily know the words of Confucius, but as a collective group that tried to appeal to US evangelical Christians, they are certainly far from the wisdom of the Bible. (Fig. 2).

For readers of this Blog, I have been quite upfront about my personal worldview and convictions. And the foregoing words of wisdom will be a well-oiled lamp that guides our path in the coming period of darkness and instability. To know what lies ahead, we need to examine what was said.

1. Strait of Hormuz Conflict - On April 20^th 2026 Trump stated: “We're going to hit them extremely hard over the next two to three weeks. We're going to bring them back to the Stone Age, where they belong." (Fig. 3). 

These words totally decimated whatever residual pro-American interests remained in Iran and hardened the resolve of Iranians to prevail over all and every Zionists’ threat.

On July 13, Trump posted on his Truth Social about a 20% American protection extortion levy on Hormuz traffic. (Fig. 4).

This was retracted the following day. How do you think the international community would react? (Fig. 5).

World opinion has already been formed by two memes, with the center of power shifted to BRICS. (Fig. 6).

I guess you all know the answer to yet another Trump threat to destroy all Iranian civilian infrastructure (Fig. 7).

The reality is that Iran’s tit for tat would be to enforce a blockade against the Red Sea by the Houthis at Bab el-Mandab, and the world will make Trump answerable for an all-critical 19% of ME oil lifeline that did not go through Hormuz. (Fig. 8).

I am therefore not concerned and will be relaxing this weekend with Mexican delights of TACO and NACHO.

2. Artificial Intelligence - I guess I am out by two weeks regarding the stock performance of Space X IPO. I gave SPCX two weeks, and it would fall below the IPO price. SPCX did that with 23 trading days. (Fig. 9).

As a one year old grandpa, I am fascinated by infant brain development andderive tremdous pleasure watching my granddaughter grow.  Inadvertently, I also became extremely skeptical of the Wall Street and Silicon Valley hype of LLMs. Input through GPU, LLMs are essentially high-speed probability pattern recognition guesswork rather than intelligence in the true human linguistic sense. The American narrative is that it got some exclusive claims on a portion of high-density chip technology, printable money to buy HBM (Ultra-High Bandwidth Memory), and therefore should dominate the AI ecosystem to the exclusion of everyone.  To my utter amazement, there is also this cult-like belief among the Mag 7 that an AI singularity will occur imminently. When that occurs, the first innovator will go through a cycles of runaway recursive self-improvement upgrades, achieve omnipotence, and then can dictate all humanity. With this obsession, immensely profitable tech titans are all spending money without caution on GPU, HBM and related R&D infrastructure to be the first achievers. I will give more details on their finances later, but first let us deal with a few known facts about human brain development, as after all, AI is fashioned after a human neural network.

(i) The Neural Network - Artificial Neural Networks (ANNs) were inspired by human brains. While human brains are complex, energy-efficient biological organs capable of continuous learning, ANNs are rigid mathematical models designed to process vast amounts of digital data to perform highly specific, singular tasks. (Fig. 10).

GPT-4 was trained simultaneously on approximately 25,000 NVIDIA A100 GPUs, while frontier models currently employ clusters of 100k+ NVIDIA Hopper or Blackwell GPUs. By comparison, the human brain has an estimated 86 billion biological neurons connected by trillions of synapses that communicate using both electrical and chemical signals. ANNs consist of a much smaller number of artificial neurons—mathematical nodes—arranged in strict layers (input, hidden, and output). The interconnections between GPU and HBM in AI language are called “parameters” and apparently the larger the number of parameters, the more powerful the “brain power” of AI. Now let’s take a look at the top of the-line LLms and compare them to a newborn baby. (Fig. 11).

A newborn baby has 5X the current top AI. Yet, despite the large number of synapses in the baby’s brain, the baby cannot tell the difference between a banana and a banana-shaped animal wastee unless the baby is trained. Bear in mind, the lessons a baby will learn is much more than visuals through a GPU but also through smell, touch, and other senses. It also depends on the human trainer, and so does AI.

(ii) Human Brain Development - From the 50 trillion synapses, a baby neural connections will grow to 1000 trillion by 3 years old before natural pruning sets in to stabilize the number of neural connections to 100 trillion. (Fig. 12).  

Synaptic pruning is the brain's way of refining itself to become more efficient. By eliminating excess or unused connections while strengthening frequently used ones, it optimizes cognitive function, reduces energy consumption, and improves the brain's ability to adapt and learn. AI although it mimics the human neural network, has no knowledge, provisions, or plans for pruning, and my expectation is that cyber weeds will grow and the drainage will clutter before humility sets in for the elites to appreciate they cannot build God in a box.

(iii) The Hard Numbers: Since 2024, the term Mag 7 has been sarcastically changed to Lag 7 because of their under performance relative to the general stock market. I have compiled a combined free cashflow chart of the hyperscalers (Microsoft, Google, Amazon, Meta and Oracle) plus Open AI, Anthropic and Space X. (AI Generated - Fig. 13).

If we just concentrate on the thick black line, the combined free cash flow (Operating cash flow minus Capex) has decidedly turned negative in 2026 and will further deteriorate in 2027. There are two separate danger signals in the horizon: -

(a) US AI service providers have found that their revenue growth is much slower than their Capex. Adoption numbers are impressive when AI is free or at a highly subsidized rate. This is a different story once the AI service providers begin to charge on a token usage basis. Enterprise users like Coinbase, DoorDash, Airbnb and Seimens immediately switched to China-based AI service providers when Chinese AI are open-sourced and charges were at a tiny fraction of what US providers charged. (Fig. 14 shows a global ranking by their most recent token usage).

(Fig. 15 shows this ranking by the cost per million token usage). A survey concluded in July 2026 showed that 80% of US AI Startups switched to Chinese Models. (Fig. 15).

(b) The second risks factor on US AI sector is their circular financing arrangement. OpenAI’s revenue grew from $2 billion in 2023 to $20 billion in 2025. Anthropic’s annualized revenue run rate surged from $87 million in January 2024 to $30 billion by April 2026, a trajectory that Salesforce took 20 years to achieve. NVIDIA’s revenue grew eightfold from $27 billion to over $216 billion between 2023 and 2026, achieving that growth in just one-third of the time it took Apple to do so during its heyday between 2007 and 2015. By nearly every reported metric, the AI market is delivering growth at a scale and speed that defies historical comparison. Yet a narrative has emerged about how much of that growth reflects genuine market demand and how much reflects the circular financing structures that have become a defining feature of AI infrastructure investment. Circular financing in AI describes investment structures where the same capital flows simultaneously as vendor payment and equity stake. A company funds its own customer’s revenue while also supplying that customer’s core infrastructure. The result is reported revenue growth that is real but not cleanly separable from investment activity. The answer matters enormously, not just for investors in AI infrastructure companies, but for every enterprise software vendor and technology buyer trying to make sense of what’s actually happening in the market. The telltale sign shows up when profit growth of these AI-related companies is derived not from operating income but other income, meaning valuation gains from their investment activities. (Fig. 16).

Their books are not clean, and Open AI has already delayed its IPO and Anthropic is planned to cash out in Oct 2026. The AI bubble when it burst will be worse than the subprime fiasco in 2008 because the excess valuation in this caper is a whopping $10 trillion. AI related stocks currently account for nearly 50% of S&P. The contagion will be ugly.(Fig. 17)

Genesis 11:1  Now the whole earth had one language and one speech.2  And it came to pass, as they journeyed from the east, that they found a plain in the land of Shinar, and they dwelt there. 3  Then they said to one another, "Come, let us make bricks and bake them thoroughly." They had brick for stone, and they had asphalt for mortar. 4  And they said, "Come, let us build ourselves a city, and a tower whose top is in the heavens; let us make a name for ourselves, lest we be scattered abroad over the face of the whole earth." 5  But the LORD came down to see the city and the tower which the sons of men had built. 6  And the LORD said, "Indeed the people are one and they all have one language, and this is what they begin to do; now nothing that they propose to do will be withheld from them. 7  Come, let Us go down and there confuse their language, that they may not understand one another's speech."8  So the LORD scattered them abroad from there over the face of all the earth, and they ceased building the city. 9  Therefore its name is called Babel, because there the LORD confused the language of all the earth; and from there the LORD scattered them abroad over the face of all the earth.