Pakistan is exporting water, not rice

The data-centre panic is aimed at the wrong target

Muzamil opens by returning to his earlier series on AI data centres arriving in Pakistan and the online backlash that they will drink the country’s water and burn its electricity. He does not dismiss the concern, but he anchors it in a number: a hypothetical one-gigawatt data centre, of the roughly thirty-billion-dollar variety Pakistan cannot yet build, would use about 0.01 percent of the country’s total water. Globally, even the projected 6.6 billion cubic metres of data-centre water use by 2027 is a fraction of Pakistan’s own 183 billion cubic metres of annual withdrawals.

“That community that was sitting and crying about the data centre,” Muzamil says, “which can supply intelligence for 0.1 percent of our water.” His point is not that data centres do not matter, they should be built well, in the right places, with the right policy, but that they are, in his phrase, an invitation the country cannot refuse. “If we run away from this,” he says, “this country will be left behind.”

Money is a contract for value, and water has value

Before he gets to policy, Muzamil detours into what turns out to be the spine of the argument: Pakistanis, he says, do not understand pricing. He gives the thought experiment of someone offering five lakh rupees a month to sit in a chair and do nothing. Most people would take it. Most people should not. If value is not being created underneath the payment, “it is a Ponzi scheme.”

He applies the same logic to a natural resource. Scarcity creates value. Value has to be priced, otherwise the resource is misallocated and wasted. He knows this will provoke his audience. “I will bring out my swords and start sharpening my knives,” he imagines them saying, “brother, you have gone mad, how can we price a blessing given by Allah?” He grants that access to water is a basic human right and must stay universal. But universal access, he insists, is a different question from correct pricing.

From water-abundant to water-scarce in one lifetime

The most sobering section is a recitation of the Falkenmark numbers. Per capita water availability in Pakistan was 5,650 cubic metres in 1951. By 2017 it had fallen to 908. Today it stands at around 660. The UN and World Bank threshold classifies anything under 1,700 as stressed, under 1,000 as scarce, and under 500 as absolute scarcity.

Pakistan, Muzamil reminds the viewer, was once one of the most water-abundant nations in Asia. “Pakistan is the cradle of the world’s oldest civilisation,” he says, “built on top of the Indus River.” Now water no longer reaches the sea in most years, the Indus delta is collapsing, and the mangroves are dying. “This is not a warning,” he says of where the number sits today. “This is literally the line after ‘you have to wake up.’”

He is visibly frustrated that this is not a national conversation. “Is there anyone in Pakistan who can save water?” he asks. He recalls going on Naeem Sikandar’s podcast in 2022 or 2023 and warning about a coming water shortage in 2025–26. The clip crossed a million views. The comments, he says, were mostly mockery, “the drawer Aristotle, the cheap Aristotle, the Chinese Aristotle.” He laughs, but the point stands.

Where the water actually goes

Muzamil then delivers the number that anchors the episode. Pakistan withdraws about 183 billion cubic metres of water each year. Of that, 94 percent goes to agriculture. Household use across 250 million people is about 5 percent. All of industry combined, cement, textiles, food, cars, bottled water, everything, is about 1 percent.

This flips the usual villain narrative. It is not the rich man washing his car. It is not the fifteen-minute shower. It is not the mineral water brand. “Pakistan’s water security threat industry,” Muzamil says bluntly, “is not that rich man.” It is a farming system that is enormous, politically protected, and technologically frozen.

He is careful to say twice that he is not attacking farmers. The next video, he tells the audience, will unpack the agricultural economy in detail. The problem is the structure around the small cultivator: no metering, no volumetric pricing, no incentive to modernise, and a handful of ultra-large landholders who benefit from keeping it that way.

The sugarcane scandal and the rice illusion

The sharpest specific example is sugarcane. Muzamil notes that sugarcane occupies about 4 percent of Pakistan’s cropland but consumes 17 percent of its water. Pakistan produces about 2 kilograms of sugar per cubic metre of water; the global benchmark is 3.5. The crop is water-hungry, the flood-irrigation methods are ancient, and the political weight of sugar barons keeps it in place. “Zardari family, Sharif family, all the big families,” Muzamil says. “They know: brother, this is a rent-seeking asset, we are easy, we get money from this without doing anything.”

He extends the same logic to rice. Pakistan exports about four million tonnes of rice a year, a fact the country takes pride in. Muzamil reframes it. “You are actually exporting water from Pakistan,” he says. Pakistan produces roughly 0.23 kilograms of rice per cubic metre of water; Egypt, another water-scarce nation, produces 0.79. The average Pakistani rice farmer uses 1,200 to 2,000 millimetres of water via flood irrigation where 800 to 1,000 would be optimal. The embedded water is subsidised, and on the world market Pakistan is undercutting competitors by giving away something it cannot spare. A single kilogram of sugar, he adds, takes about 1,750 litres of water to produce.

Corporate farming for the Gulf is water leaving the country

Muzamil turns to the corporate-farming deals with Gulf investors. He references retired Major General Tahir Aslam telling Arab News that 40 percent of the Green Pakistan Initiative’s output will be exported to Gulf countries, with a first order of $25 million. He is not against foreign investment in principle. His objection is that without proper water pricing, “you are basically exporting water. You just don’t want to call it water.”

He draws the parallel with bottled water: if a company started bottling 10 percent of Pakistan’s water and shipping it to the Middle East, “you will be up in arms.” Doing the same thing through wheat, fodder and rice is politically invisible. He is equally blunt about how these decisions get made, comparing them to the IPP contracts of the 1990s, long-lived commitments signed by people who “are not experts” in the underlying economics, whose consequences arrive twenty years later.

Why free water guarantees waste

Muzamil returns to why pricing matters at a behavioural level. Electricity in Pakistan is expensive, so people switch off lights and fans. “Every button in your house, you press it yourself,” he says. Water is effectively free at the tap and at the tubewell, so nobody thinks about it. Farmers flood their fields because there is no cost signal telling them not to.

He shares an anecdote from his World Bank work between 2012 and 2014 on the Water and Sanitation Programme, dealing with WASA authorities in Faisalabad, Lahore, Multan, Peshawar, Karachi and Rawalpindi. He saw societies next to canals with no piped water at all, residents pulling muddy water out of the canal with a jury-rigged motor, decanting it into blue containers, and paying donkey-cart operators to deliver it. In DHA Karachi, tanker deliveries cost between ten and twenty-five thousand rupees because the pipe network is either missing or sabotaged. The “tanker mafia,” he says, physically tears up new lines because their diesel-fuelled business depends on the pipes never working.

His own water bill in a central Islamabad F-sector house, he estimates, comes to around ten to twelve thousand rupees a year. He thinks a house of that value should be paying at least 150,000. His US water bill, he mentions, runs $100–150 a month in an area with abundant rainfall. He is not suggesting Pakistanis pay American prices. He is suggesting they pay something meaningful enough to change behaviour.

Volumetric pricing, not more dams

By the end, Muzamil narrows his recommendation to one sentence. Water pricing in Pakistan is currently done on a flat per-acre basis, which means a farmer using ten litres and a farmer using ten crore litres pay the same. That has to become volumetric. “You have to pay per volume of water that you extract.” The IMF’s 2015 report says the same thing, and notes that provincial water authorities recover only 25 percent of the running cost of the canal system, with the rest subsidised out of general taxation, meaning water is not really free, it is just paid for by taxpayers in a way that hides its cost. Provinces already sink around 8 billion rupees a year into keeping the existing canals barely operational.

He anticipates the political attack that pricing water is regressive. His answer: ordinary households consume so little of the total that even a properly priced system would leave 250 million people paying almost nothing more. The real bill would land on large landholders who currently flood their fields. Once they had to pay, drip irrigation, tunnel farming and modern methods would follow, not out of virtue, but out of cost pressure. He is dismissive of supply-side-only fixes: “you build a dam, you build ten more things,” he says, but until pricing changes, demand keeps outrunning supply.

What comes next

Muzamil is honest that he does not expect this video to do the numbers his AI content does. “People don’t really care” about water, health and education topics, he says. “People just, once again, teach them how to double money with money.” But he wants the framing in the audience’s head: data centres are a rounding error, sugarcane is 17 percent, rice exports are embedded water, corporate farming for the Gulf is water leaving the country at a subsidised rate.

He closes by previewing the next episode on the agricultural economy, and two upcoming series, one on the semiconductor supply chain based on a conversation with a Pakistani engineer at ASML, and one on where the real opportunities in AI will sit over the next five years. On the AI-jobs panic, he is unambiguous: “I 100 percent reject the narrative that jobs will end.” But that is for another episode. This one was about the water Pakistanis are not counting.