The Gulf's Paradox

March 2026. As conflicts escalated across the Middle East, the world's eyes turned to missiles and fighter jets. But on the quieter front of this war, there was a very different target: water.

The Gulf region sits on some of the world's most enormous oil and gas reserves — wealth that shapes the global economy. Yet beneath that same surface lies an equally serious absence: fresh water. The world's richest oil nations are, at the same time, among its most water-scarce.

Water Scarcity

Annual rainfall across the Gulf averages between 70 and 130 millimeters — and in some areas, it barely rains at all. According to NASA, the Middle East has been experiencing its worst drought in 900 years since 1998. Seven of the world's ten most water-stressed countries are in this region, where annual water withdrawals reach up to eight times the available renewable supply. The UN World Water Development Report 2024 found that 83% of the region's population lives under extremely high water stress. Bahrain, Kuwait, Oman, and Qatar rank among the most critical cases.

Desalination: A Vital Solution, a Fragile System

Faced with these conditions, Gulf states have been turning to desalination — converting seawater into drinking water — since 1938. Today, 90% of Kuwait's drinking water, 86% of Oman's, and 70% of Saudi Arabia's comes from these plants. The Gulf alone accounts for roughly 40% of the world's desalinated water, with over 400 facilities operating along its coastline. But the system is deeply fragile: entire cities depend on just a handful of large plants. Most of them run alongside power stations — meaning an attack on the energy grid can knock out water production just as quickly. This vulnerability isn't new. During the 1990–1991 Gulf War, Iraqi forces deliberately destroyed much of Kuwait's desalination capacity. That precedent has come back into sharp focus.

Attacks: A Region Under Threat

The March 2026 conflict struck water infrastructure directly. The US launched missiles from the Jufair base in Bahrain, hitting a desalination plant on Iran's Qeshm Island at the entrance to the Strait of Hormuz. According to Iran's Foreign Ministry, the strike disrupted water supply to 30 villages, forcing thousands to turn to unsafe water sources and significantly raising the risk of waterborne illness in the area. Kaveh Madani, Director of the UN University's Institute for Water, Environment and Health, called the attack "a wake-up call for the region."

But Qeshm was not the only target. In the same period, Iran was reported to have damaged a desalination plant in Bahrain with a drone. Missile debris landed near a power and water complex in Fujairah, UAE. A strike on Dubai's Jebel Ali port came dangerously close to the city's 43 desalination units, which produce over 600 million cubic meters of water annually. And in Kuwait, debris from an intercepted drone caused a fire at the Doha West Power and Water Station.

Analysts point out that none of this is coincidental: water facilities are fixed, coastal, high-value targets — strategically hard to resist. With Iran reportedly capable of producing around 10,000 drones per month, the scale of the threat becomes even clearer.

These attacks also raise serious questions under international law. The Geneva Conventions explicitly prohibit attacks on water facilities essential to civilian survival.

Regional and Global Implications

The impact of these strikes extends well beyond the region. Iran's closure of the Strait of Hormuz disrupted roughly 20% of global oil trade, forcing tankers to reroute around Africa. Since about a third of the world's fertilizer trade also passes through the strait, global food prices have come under significant pressure. Experts warn that the real crisis may still be coming — in the form of acute water shortages over the months ahead. Ed Cullinane of Global Water Intelligence puts it plainly: with ongoing strikes, an economic crisis, and a serious water shortage all converging, it is hard to even imagine how this summer will unfold.

When a new park opens in a city, it is almost always seen as a positive development. More trees, more space, cleaner air — it sounds like something no one would argue against. And in most planning discussions, green spaces are treated exactly this way: as simple, universally beneficial interventions.

But the reality is more complicated.

Research consistently shows that being in natural environments improves individual well-being. People feel less stressed, mentally more refreshed, and generally better after spending time in green spaces. These effects are well established. What is less clear — and often taken for granted — is whether these individual benefits translate into stronger communities.

They don't. At least, not automatically.

Just because a park exists nearby does not mean people will trust each other more, feel a stronger sense of belonging, or build meaningful social connections. These outcomes depend on other factors: how the space is designed, whether it feels safe, who uses it, and whether people actually want to spend time there. In many cases, what matters is not how large a green space is, but how it is experienced. A small, well-maintained, thoughtfully designed space can offer more than a large but neglected one.

From an environmental engineering perspective, this distinction feels familiar. It is not enough for something to exist as infrastructure; it has to function properly within a system. Water, for example, is not useful simply because it is present somewhere. It needs to be accessible, evenly distributed, and continuously maintained. Green spaces operate in a similar way. Their physical presence does not guarantee their social function.

And in some cases, the effects go in the opposite direction.

One of the more uncomfortable findings in recent studies is the phenomenon of green gentrification. When new parks or green investments make an area more attractive, housing prices tend to rise. Over time, this can lead to the displacement of long-term residents, particularly lower-income communities. The process is often gradual and difficult to trace directly, but the outcome is clear: the people who were supposed to benefit from these improvements may no longer be there to experience them.

What is lost in this process is not just physical proximity to green space, but something less visible — social ties, familiarity, and a sense of belonging built over years. These losses rarely show up in data, but they shape how communities function.

None of this means that green spaces are not valuable. They clearly are. But their benefits are not automatic, and they are not equally distributed. Whether a green space supports well-being and social cohesion depends on decisions made long before it is built — decisions about design, accessibility, maintenance, and who is included in the planning process.

Treating green spaces as neutral or universally positive overlooks these dynamics. They are not just environmental features; they are part of a broader social and political system.

So the question is no longer whether cities need more green space. That part is obvious.

The more important question is who that green space is actually for.

At some point, I realized something: projects that once required teams were now moving faster, with fewer people. Work that used to depend on groups was turning into processes. Eventually, a question became unavoidable:

Where do I stand in this?

The Myth of Safety

For a long time, there was a common belief: AI handles repetitive work, but science is different. It analyzes, interprets, discovers. Researchers would always be needed.

Then came AlphaFold.

Predicting protein structures had been one of biology's hardest problems. It required years of lab work, expensive equipment, and specialized teams. AlphaFold reduced that process to hours — and produced results at a scale no human team could match.

It was a breakthrough. And at the same time, a quiet reduction of roles.

Crystallography workloads dropped. Parts of bioinformatics became automated. Computational modeling teams began to shrink — not disappearing, but consolidating. From the outside, it didn't look like disruption. It wasn't a headline.

But the reality was simple: fewer people were needed to do the same work.

And this isn't limited to biology. Environmental modeling, climate analysis, molecular simulations, materials science — anywhere work is driven by large datasets and pattern recognition, AI is already embedded in the system.

Where the Pressure Is Highest

Not all roles are affected equally.

Data analysts are among the most visibly impacted. Data cleaning, structuring, processing — once daily tasks — can now be done in minutes. Computational modelers face similar pressure. Simulations that once required deep expertise are increasingly handled with AI-assisted tools. Literature reviewers are also affected — systematic reviews that used to take months are now significantly faster.

The common thread is clear: if a role is repetitive, pattern-based, and data-intensive — AI is already inside it.

This doesn't mean these roles will disappear. It means fewer people will be needed to do them.

Less Exposed — For Now

Some roles are more resistant, at least for now. Field researchers. Experimental scientists. Roles that require direct interaction with complex physical environments. AI still struggles with the physical world, uncertainty, and human interaction.

But "for now" matters.

From the Inside

I'm not observing this from the outside. I'm inside it.

I've been working in LLM evaluation — analyzing outputs, identifying errors, assessing quality. For a while, the work felt stable. Then something shifted. Parts of the evaluation process started becoming automated. Tasks I used to do manually were now handled by other models.

This wasn't surprising. The speed was.

At that point, the real question wasn't whether AI would affect my work. It was how I would position myself within that change.

I started moving toward areas like RAG systems, agent architectures, and fine-tuning. Not because they are "safe," but because they sit closer to where decisions are made — not just where outputs are evaluated.

In AI, there is no safe place. There is only positioning.

What Actually Matters

"Learn AI and you'll be fine" is too simplistic.

The real question is this: which part of your work is pattern recognition, and which part is judgment?

AI is extremely good at replicating patterns. Judgment is still harder to replace.

Domain expertise still matters — but not on its own. It needs to be combined with technical understanding and critical thinking. For those who can build that combination, the field isn't shrinking.

It's transforming.

The Real Shift

The lab is changing. Not because science is disappearing — but because fewer people are needed to do the same work.

The question is no longer whether AI will affect science.

It's where you will stand when it does.