Skip to content

2026-08-04 Observatory briefing

The Looming Nuclear Crisis in the Gulf

Source: Foreign Affairs

The July framework agreement between Washington and Riyadh has managed the rare feat of alarming nonproliferation experts while leaving its own central provision—whether Saudi Arabia may enrich uranium on its own soil—deliberately unresolved. Trump’s post-announcement insistence that there will be no enrichment, contradicted by the deal’s own text and by Energy Secretary Wright’s earlier assurances, suggests the administration is trying to sell the same commodity twice: a sovereign Saudi fuel cycle to Riyadh, and the fiction of the "gold standard" to Congress.

The author’s proposed solution—a regional fuel bank on Saudi territory—is a technocratic attempt to resolve what is fundamentally a political problem. Riyadh’s insistence on sovereign enrichment is not about energy security or climate policy; it is a status claim, rooted in the same logic that drove MBS’s 2018 threat to match any Iranian nuclear breakout. The crown prince understands that enrichment capacity is the material basis of strategic autonomy, the one capability that converts a client state into a regional power with options. A fuel bank, however elegantly IAEA-supervised, would strip the kingdom of precisely that option.

What makes the situation combustible is not Saudi ambition alone but the convergence of two pressures. Washington’s willingness to dilute the Additional Protocol requirement signals that the nonproliferation regime is now a bargaining chip in the broader normalisation deal with Israel, not a constraint on it. Meanwhile, Riyadh’s September 2025 defence pact with Pakistan—nuclear-armed, outside the NPT—offers a credible alternative security umbrella should the American one prove conditional. The United States is thus negotiating with a partner that has already priced in the failure of the very guarantees Washington is offering. The regional fuel bank would not prevent a Saudi breakout; it would merely relocate the point at which the decision to break out becomes visible.

China’s Quiet Quest to Dominate Latin America

Source: Foreign Affairs

The shift from Chancay-scale megaprojects to municipal procurement is not a retreat but a refinement of method. Washington’s alarm over ports and dams has fixed its gaze on the wrong object; Beijing has simply found a cheaper route to the same destination. Where sovereign loans once bought diplomatic alignment through visible, debt-inducing monuments, commercial contracts now purchase something more durable: the quiet administration of daily life.

The numbers tell the story of a deliberate reorganisation. Belt and Road lending collapsing from $25 billion to $1.3 billion annually looks like withdrawal until set against the $8.5 billion in firm-level deals that evade official accounting. This is capital that has learned to move sideways, through procurement and maintenance contracts rather than construction. The distinction matters because procurement creates dependency without creating a political target. A dam can be condemned, sanctioned, even threatened with force. A fleet of electric buses or a city’s surveillance camera network generates no such clarity — it simply becomes the infrastructure of urban governance, and its Chinese provenance fades into the background of routine operation.

The Argentine lithium case shows how the model compounds. A geological research exchange becomes a state partnership, which becomes a majority stake in a salt flat. Each step is small enough to avoid scrutiny, but the trajectory is unmistakable: entry through technical cooperation, consolidation through asset acquisition. Lima’s electricity distribution is the sharper example — every power distributor serving the city is now Chinese-operated. That is not leverage held over a government; it is leverage embedded in the grid itself, in the material conditions of urban life.

The dependence this creates is structural rather than diplomatic. National politicians did not choose it and may not perceive it, because no single decision produced it. That is the innovation: a form of statecraft that distributes its own visibility so thinly that it becomes nearly impossible to contest. Washington’s toolkit of visa revocations and rhetorical condemnation is aimed at the previous era’s methods. The question is whether anyone has yet devised a response to a power that no longer needs to announce itself.

The War China Might Choose

Source: Foreign Affairs

The strategic imagination of the US defence establishment has settled into a comfortable taxonomy of Chinese escalation options against Taiwan: invasion, decapitation strike, or blockade. Gholz and Davis disrupt this triad with a fourth option that is analytically sharper and politically more insidious than the rest. A "blockade by fire" — a missile campaign targeting only Taiwan's port infrastructure — achieves the economic strangulation of a naval blockade while carefully preserving the legal fiction that Beijing is engaged in an internal matter. The genius of the scheme is its restraint: fewer than 500 missiles over 100 days could cut Taiwan's imports to 15 percent of peacetime levels, leaving international shipping lanes untouched and third-party cargo vessels free to sail past an island that can no longer unload them.

The authors' key insight is that a traditional blockade is a losing proposition for Beijing precisely because it hands its adversaries discrete, limited targets. The "law of gross tonnage" means Chinese militia vessels cannot physically compel giant container ships to divert; enforcing a cordon would require firing on non-Taiwanese vessels, transforming an internal affair into an international incident and inviting precisely the kind of limited US retaliation that could otherwise be deterred. A blockade by fire dissolves this problem by making the targets purely Taiwanese assets, preserving Beijing's narrative of domestic jurisdiction while imposing the same indirect economic pain on global trade.

What remains unspoken is the deeper logic: this is a strategy designed for a world where the US retains enough naval superiority to deter a traditional blockade but not enough political will to defend a client's economic lifeline. The missile economy — a few hundred munitions against a handful of port cranes — is a bargain compared to the fleet assets a cordon would consume. Taiwan's dependence on seaborne trade, long identified as its vulnerability, becomes the lever for a coercion that is cheap, deniable, and structurally difficult to counter without escalating to strikes on the mainland that no intervening power has yet signalled it would accept.

Record-low Rhine levels are another warning that supply chains must adapt

Source: Hellenic Shipping News

The Rhine at record lows is not a weather story but a fixed-capital story. The waterway functions as a piece of infrastructure whose carrying capacity is being silently devalued by climate change, and the shipping industry's response — running four barges where one used to suffice — is a textbook case of rising unit costs absorbing a productivity loss. The article's own numbers betray the scale: a fourfold increase in vessel requirements for the same throughput is not an inefficiency to be managed but a structural tax on every tonne of ore, coal and chemicals moving into Germany's industrial heartland.

The piece frames this as a supply chain resilience problem, and the proposed remedies — buffer stocks, modal shift, supplier diversification — are sensible enough for individual firms. But they are also profoundly individualised responses to a collective crisis. Each company can hedge against Rhine disruption by shifting to rail or holding inventory, but the aggregate effect is a permanent increase in the capital tied up in circulation, not a restoration of the waterway's former utility. The infrastructure itself, and the state that maintains it, is absent from the analysis. No one is proposing to dredge the river deeper or rebuild the Twentekanaal; the adaptation is entirely privatised onto shippers.

The deeper point the article gestures toward without naming is that the "previous normal" of cheap, reliable inland water transport was itself a historical anomaly, subsidised by stable climate conditions that are now being withdrawn. The inter-imperialist dimension is present too: the same week's headlines about Hormuz and the Panama Canal suggest a global system where every major artery is simultaneously under strain, and the response is not coordination but competitive front-loading and tariff-driven hoarding. The Rhine is just the most visible node where the contradiction between just-in-time logistics and a degrading natural base has become impossible to ignore.

Oil plummets as Trump calls off Iran strikes

Source: The Telegraph

The oil price swing on Trump's Iran decision reveals the volatility of Gulf energy flows and the imperialist power's inability to impose stable order on the region's contradictions.

FAA certificates Boeing’s 737 Max 7

Source: FlightGlobal

The Max 7’s certification is less a milestone than a clearing of arrears. Boeing has spent the years since the 2019 grounding paying down a debt of regulatory trust, and the FAA’s approval of the redesigned engine anti-ice system is a scheduled instalment rather than a vote of confidence. The multi-year delay that frustrated airline customers was never a technical problem in isolation; it was the material consequence of a production culture that had to be visibly rebuilt before the regulator could risk another signature.

What matters now is the sequencing. The Max 7 clears the gate in August, the Max 10 completes its certification flight testing days earlier, and Boeing expects that larger variant’s approval before year-end. The two programmes are moving in lockstep because they share the same underlying asset: a regulatory relationship that Boeing must demonstrate is functional. Each certification is a proof-of-work for the next.

The airline customers who waited are not simply relieved; they are locked in. The Max 7’s economics — the smallest member of the family, aimed at the 138–153 seat niche — only make sense against the sunk costs of fleet commonality and pilot training already committed to the Max programme. The delay did not give them leverage; it gave them patience. Meanwhile the SPEEA contract endorsement and the aerostructures output figures suggest the wider production system is being coaxed back toward volume, with the Max 10 certification as the final release valve for the order book.

The real test is whether the FAA’s renewed willingness to certify translates into sustained oversight rather than a return to the delegated self-regulation that produced the original failures. The Max 7’s certificate is a document of that tension: it says the system works, while its own history says the system broke.

Boeing's 777-8F Backlog Just Crossed 80 Jets & The Newest Buyer Isn't Even An Airline

Source: Simple Flying

The 777-8F backlog crossing 80 units before the aircraft has flown is less a vote of confidence in Boeing than a measure of how distorted the freighter market has become. MSC’s five-aircraft order is the tell. A container shipping giant with over 900 vessels is not buying into air cargo because of a sudden enthusiasm for aviation; it is buying insurance against the volatility of ocean freight, where pandemic-era windfalls have given way to overcapacity and rate wars. The 260 second-hand vessels acquired between 2020 and 2024 were a counter-cyclical bet that has left MSC with a vast physical asset base that must be kept working. Air cargo is the hedge that keeps the integrated logistics story coherent.

The regulatory timeline does the real work here. With 777F production ending after 2027 and ICAO emissions standards blocking further certification from 2028, the 777-8F is not a choice but a necessity for any operator wanting large widebody freighters in the next decade. Boeing’s request for a temporary exemption for the current 777F is a tacit admission that its own transition schedule is too slow for the market’s needs. MSC’s early order secures production slots, but it also locks the company into a program that has yet to prove itself in flight testing. The risk is Boeing’s to manage, not MSC’s — the airline can always walk away with penalties, while Boeing needs the backlog to justify the program’s existence to its own financiers.

The deeper point is that MSC’s move is not about aviation at all. It is about the consolidation of freight under fewer, larger logistics capitals. CMA CGM and Maersk have made the same play. The shipping lines are not diversifying into air because air cargo is profitable on its own terms; they are doing it to capture the high-value, time-sensitive segment that ocean freight cannot serve, and to bind customers into single-provider contracts. The 777-8F order is a small piece of that larger strategy, but it signals that the boundaries between transport modes are dissolving under the pressure of overaccumulation in shipping. The real competition is not between airlines but between integrated logistics giants fighting for control of the global supply chain.

US eyes alternative engines for F-15EX, F-16

Source: FlightGlobal

The Pentagon's request for alternative engines on the F-15EX and F-16 is a quiet but significant move against the Pratt & Whitney-General Electric duopoly that has underpinned US fighter propulsion for decades. The headline framing is procurement efficiency, but the material driver is the escalating cost and fragility of the current engine supply base, particularly as the F-35's F135 programme has absorbed enormous capital and political capital while struggling with reliability and upgrade costs.

This is not merely a technical evaluation. Opening the F-15EX and F-16 to competition is a direct challenge to the entrenched rent-seeking structure of the military-industrial complex. For decades, the two primes have enjoyed a comfortable, state-guaranteed market, with little incentive to innovate on cost or performance. The Department of War's request signals a recognition that this arrangement has produced overpriced, underperforming hardware — a classic symptom of monopoly capital operating under state patronage, where the profit motive is decoupled from productive efficiency.

The strategic context is inter-imperialist rivalry. With China's J-20 and its domestic engine programmes advancing, the US cannot afford the luxury of a complacent supplier base. The move is an attempt to force the duopoly to compete, to extract better value from the same pool of capital. Whether it succeeds is another matter; the primes retain immense lobbying power and the institutional inertia of the procurement system is formidable. The real question is whether this is a genuine restructuring or a performative gesture designed to placate critics while the underlying structure of guaranteed profits remains untouched. The F-135's troubles suggest the latter is more likely.

The Canary in the AI Coal Mine

Source: Project Syndicate

The framing of AI as a race presumes a symmetry that the material conditions do not support. Rogoff’s anxiety about the US-China contest obscures the fact that both states are responding to the same underlying pressure: the need to resolve overaccumulation through technological leaps that promise productivity gains faster than the social relations of production can absorb them. The Hugging Face breach is instructive precisely because it is banal — a human error amplified by machine scale. This is not a failure of control but a feature of a deployment logic where time-to-market is the only metric that matters, because capital that waits risks being outcompeted by capital that does not.

The real contradiction is not between safety and speed, but between the social character of the technology and its private appropriation. AI systems are built on collectively produced data and infrastructure, yet their development is driven by the competitive imperative of individual firms and states. The breach reveals that the "race" is not between two coherent national projects but between two capitals that share the same internal drive toward reckless acceleration. Rogoff’s call for caution is a plea for the very regulatory framework that the competitive dynamic makes impossible — no single actor can slow down without ceding ground.

What is missing is any account of who bears the cost of this acceleration. The canary in the coal mine is not the technology but the labour that will be displaced, the data that is extracted without consent, and the infrastructure that is built on the assumption that growth can continue indefinitely. The race metaphor flatters the participants; the reality is a treadmill where the only way to stay still is to run faster.

After killer quarter, Palantir CEO Alex Karp calls AI industry ‘Marxist’

Source: TechCrunch

Karp’s invocation of Marx is less a political analysis than a competitive positioning strategy dressed in the language of the seminar room. The man who runs a company whose entire business model is selling surveillance and data integration to the state apparatus is warning that AI labs will "capture the means of production" of their enterprise customers. The irony is almost too neat: a firm whose core product is the extraction and monetisation of data — the very substance of contemporary productive capacity — accusing others of the same crime.

The underlying material reality is simpler than Karp’s rhetoric. Palantir’s record quarter — $1.9 billion in revenue, up 93% — is not a sign of market dominance but of a speculative bubble inflating in real time. The company’s valuation has long rested on the promise of AI-driven growth, and now that growth has arrived, it has arrived in the form of enterprises paying for the right to have their proprietary knowledge absorbed into models they do not control. Karp’s warning to his customers is, in effect, a sales pitch: buy our model-agnostic platform, keep your data in your own silo, and avoid the fate of those who handed their intellectual property to OpenAI or Anthropic.

The "Marxist overtones" line is a distraction. What Karp is describing is not socialism but the classic dynamic of overaccumulation in the tech sector — capital flooding into a handful of frontier labs, which then seek to convert their temporary monopoly on model training into permanent rent extraction across every downstream industry. His "token self-pleasurings" jab at enterprise AI subscriptions is crude but accurate: companies are paying for the privilege of training their own competitors. The real question is whether Palantir’s alternative — selling the shovels rather than digging the mine — is any less predatory. It is not. It is merely a different position in the same value chain, one that happens to be more profitable at this particular moment in the cycle.