The mistake Bhringi makes is subtle: Shiva is real. His devotion is not directed at something false. His error is believing that because one principle is profound, everything complementary to it can be discarded.
That distinction matters enormously for AI.
Bhringi as the AI market
Imagine Shiva = computation / intelligence / the model and Shakti = everything required for that intelligence to exist economically in the real world:
- electricity and data centres
- chips and supply chains
- capital and interest rates
- customers willing to pay
- human labour and institutions
- regulation, trust and legitimacy
- ultimately, cash flow
The market becomes Bhringi when it says:
“The model is everything.”
More parameters. More GPUs. More agents. More compute. Bigger valuations.
Meanwhile Shakti is standing right beside Shiva saying:
“Yes—but who powers it, who pays for it, and what productive activity does it actually create?”
That isn't an abstract constraint anymore. Hyperscaler capital expenditure is estimated around $795 billion in 2026 and more than $1 trillion in 2027, while AI-related optimism continues to support major equity indices. At the same time, the physical buildout is pushing directly into electricity generation: a utility announced just this week that it plans $1.8 billion of investment partly to serve a proposed Google data centre.
Ardhanarishvara is what good leadership sees
Shiva eventually appears as Ardhanarishvara precisely to destroy Bhringi's false separation.
For a CEO or investor, Ardhanarishvara is the integrated balance sheet:
technology + economics
intelligence + energy
automation + customers
software + physical reality
vision + constraint
Exceptional leaders don't merely identify the technological breakthrough.
They understand the whole coupled system around it.
That's why leadership becomes dangerous during hype cycles. The organisation begins rewarding people for demonstrating greater devotion to the narrative rather than for identifying constraints.
The board asks:
> “What's our AI strategy?”
instead of:
> “What expensive human or computational process can we now replace, and what measurable surplus does that create?”
Those are radically different questions.
And then Shakti withdraws
This is the brutal part of the Bhringi metaphor.
Parvati doesn't have to defeat Bhringi.
She merely withdraws what was always sustaining him.
And suddenly only the skeleton remains.
Markets work similarly.
You don't necessarily need AI technology to fail for an AI boom to unwind.
The models could become dramatically better while some AI investments still become terrible investments.
All that has to happen is for one or more complementary inputs to stop cooperating:
capital becomes expensive, electricity becomes constrained, utilisation disappoints, customer willingness to pay plateaus, depreciation outruns revenue, or expected productivity fails to arrive quickly enough.
That distinction is especially relevant now. AI-related equities have remained remarkably strong even while enormous infrastructure requirements and higher borrowing costs have caused investors to question the economics of the buildout.
And this gives us perhaps the most useful interpretation of AI hype:
> Hype doesn't require the underlying technology to be fake.
It occurs when the reality of the technology is used to justify ignoring the economics surrounding it.
Railways were real.
The internet was real.
Fibre optics were real.
AI can be real too.
Price is a completely separate question.
Then comes Shiva's third leg
The strangest part of the myth might actually be the most relevant to markets.
Bhringi has reduced himself to a skeleton and can no longer stand.
Shiva gives him a third leg.
Think about what happens during a gigantic capital cycle when economics starts becoming uncomfortable.
More financing appears.
Strategic investors step in.
Suppliers finance customers.
Debt structures become creative.
Governments subsidise infrastructure.
Profitable divisions cross-subsidise speculative ones.
And extremely wealthy companies can keep spending long after a normal company would have been forced to demonstrate returns.
We can already see unusual financing surrounding the AI buildout: SoftBank recently raised $11.1 billion in high-yield bonds while pursuing huge AI commitments, and Nvidia has been involved in efforts to mobilise financing around AI infrastructure buyers.
Those mechanisms don't automatically mean the investments are bad.
But they can allow Bhringi to remain standing for much longer than observers expect.
And that produces a dangerous market inference:
“It is still standing, therefore the economics must work.”
Not necessarily.
It may simply have acquired another leg.
---
There is a powerful leadership lesson buried underneath all of this:
The weak leader worships the dominant narrative.
The contrarian leader worships the opposite narrative.
The great leader sees the coupled system.
So the interesting question isn't:
“Is AI a bubble?”
Nor:
“Will AI change everything?”
Both can be intellectually lazy questions.
The better question is:
> Where is Shiva being priced while Shakti is being ignored?
Because AI capability can continue advancing enormously while capital discovers that intelligence is not synonymous with economic value.
And when markets finally rediscover the difference, it isn't necessarily Shiva that disappears.
Shakti simply presents the bill.
#AI #ArtificialIntelligence #Markets #Leadership #Technology #Economics #Strategy #Infrastructure