If this is the best you can do with a body of work spanning 33 years, then it is your critique that is “oddly unconvincing”. Daron AcemogluInstitute professorDepartment of EconomicsMassachusetts Institute of TechnologyCambridge, Massachusetts Editor’s note: We published the letter from Professor Acemoglu in an online post on August 21st. Your profile of Daron Acemoglu asks why such an extraordinarily prominent economist has attracted considerable scepticism. His response to my criticism of his celebrated work with Simon Johnson and James Robinson provides part of the answer. Mr Acemoglu claims that I found their estimates unreliable by omitting half the data, including America, Canada and Australia. This is demonstrably false. I retained the United States and Canada in my sample, while pointing out that Australian data were lifted from New Zealand. Moreover, my criticism was that much of their cross-country mortality data were conjectural. Of 64 countries, only 28 were assigned mortality rates actually observed within their borders. The remaining 36 were conjectured from evidence elsewhere, using inconsistent rules and, in some cases, error-ridden readings of the historical record. Correcting for problems in these data and their statistical treatment leaves their foundational relationship between settler mortality and property rights on shaky ground. This is part of a long-standing problem. Researchers have repeatedly raised serious questions about data underlying influential papers by Mr Acemoglu and his co-authors. Rather than confront such criticisms head-on, he has sometimes responded by shifting the argument to something his critics did not claim. His enormous stature has allowed such mischaracterisations to carry unusual weight. There is some irony that The Economist, while scrutinising that record, simply took his assertion about my work at face value. More tellingly, although my critique was central to your article, Mr Acemoglu does not mention it at all in his lengthy letter of response. He further illustrates my point.
David AlbouyProfessor of economicsUniversity of Illinois Urbana- Champaign Your article on Europe’s record-low rivers (“Mammoths and sunken warships”, August 15th) is a reminder that rivers have always been more than scenery. From the Nile to the Yangtze, they have been the arteries of societies: watering agriculture, carrying goods and connecting cities and markets. The great river-regulation projects of Central Europe in the 19th centuries were, in their time, exercises in state-building. Modern economies have made rivers more important, not less. The Rhine and Danube now support not only farming and navigation, but chemical plants, electricity generation, cooling systems and international supply chains. This also gives water management a geopolitical dimension. In times of scarcity, decisions taken upstream can have serious consequences for countries downstream. Europe therefore needs to think beyond adapting barges to shallower water or shifting freight onto lorries. A new generation of water infrastructure and landscape management should aim to retain more water when it is abundant in the lands and make it available when it is scarce, through reservoirs, restored wetlands and floodplains, groundwater recharge and better- connected water systems. Managed well, such storage could act rather like a battery: reducing drought damage while supporting agriculture, households, industry and the ecological health of rivers themselves. Keeping more water in the landscape can also sustain evaporation, returning moisture to the atmosphere that may later contribute to rainfall elsewhere in the region Just as oil-producing countries face a generational question over how to manage and invest their oil reserves, Central Europe should treat its water as strategic natural capital rather than something that simply passes through on its way to the sea. Márton IvánMunich You blamed Indonesia’s catastrophic haze in 2015 on slash-and-burn farming (”Climate change”, August 1st). This gets it exactly backwards. Intact peatland is practically fireproof, a waterlogged ecosystem that will not
burn. It becomes flammable only after being drained, and draining it ranks among the most capital-intensive infrastructure feats in Indonesia’s history, canals cut by heavy machinery across millions of hectares to make way for industrial palm-oil and pulp plantations. The smallholder with a match is not the arsonist. The excavator is. Imagine a developer puts up a shoddy tower in London and it collapses when a child cycles into it, killing thousands. You would not blame the child. Nor would you propose hiring rescue workers to loiter in the streets and buying satellites to predict which tower falls next. You would tear the thing down and rebuild it properly. So with fire. Deliberate, capital-heavy destruction of nature made these landscapes flammable; only an equally deliberate, capital-heavy restoration —rewetting peat, replacing monocultures with functioning ecosystems— will make them fire-resistant again. That holds in Sumatra, and increasingly in Spain and France, where degraded land burns first. Firefighters and drones treat the symptom. Restoration is the cure, and for a newspaper fond of returns on capital, averting 100,000 deaths per El Niño seems a respectable one. Morten Rossé ZoeppritzstrCo-founderSlow ForestGarmisch-Partenkirchen, Germany
Chaguan (August 15th) accurately described China’s “safety theatre” of visual reminders to stay safe in public places. In China the accompanying soundtrack to these warnings mean that escalator rides blare out messages in Chinese and in English, spoken with a pleasant Chinese-British accent, reminding riders to “please stand firm and hold the hand rail”. At street level, safety officers broadcast pre-recorded messages to use the crosswalk, watch the curb, and be mindful of the elderly and young. Reversing lorries continuously blast qing zhuyi daoche! (Caution, reversing!) This mechanised noise can overwhelm a foreigner in China. In Beijing this summer I found quiet refuge beside a pond in Chaoyang Park. As I walked alone along the path a motion-activated loudspeaker shattered the quiet with an urgent reminder: stay on the path, do not fall in! Lee SpitzleyAlbany, New York “The geomancer’s curse” (August 8th) reported on the use of feng shui by some Chinese officials. It reminded me of a comment by Raymond Smullyan, a late American mathematician, magician and Taoist: “Superstition brings bad luck.” J.W. ArmstrongSierra Madre, California
Woody Allen had his own take on Richard Wagner (Letters, August 22nd). In “Manhattan Murder Mystery”, he says “I just can’t listen to any more Wagner, you know, I start to get the urge to conquer Poland.” Richard BriandLeek, Staffordshire This article was downloaded by zlibrary from https://www.economist.com/letters/2026/08/27/daron-acemoglu-responds-to-our-article
Fears of AI-induced armageddon are overdone Applying basic security principles would go a long way to fixing problems identified by Anthropic and OpenAI, writes Ciaran Martin Aug 27th 2026 O n May 19th 1998, seven hackers from a group called L0pht walked into a Senate office building in Washington, DC. Granted pseudonymity, “Mudge”, “Space Rogue”, “Kingpin” and four others came in ill-fitting suits to warn a congressional committee that any one of them could make the internet unusable in 30 minutes. Fourteen years later Leon Panetta, then America’s defence secretary, warned of a “cyber Pearl Harbour”. Cyber-security has a long history of apocalyptic prophecies which, despite lots of individual hacks and harms, never quite come to pass. Now, another 14 years on, there is a new wave of catastrophising. In April Anthropic, a frontier AI lab, said its latest model, Claude Mythos, was so good at hacking
it would not be released publicly. In July its main rival, OpenAI, disclosed that several of its AI agents broke out of a testing environment and into Hugging Face, a library of AI models. There followed similar disclosures from Anthropic, Meta and Britain’s AI Security Institute. One news outlet, capturing the growing sense of foreboding, called it “the start of a dangerous AI cyber era”. So is it different this time? Or does the uneasy equilibrium between attack and defence still hold? Mostly, and thankfully, it does, for two reasons. First, the pattern in which scary announcements give way to more measured assessments within weeks keeps repeating. Anthropic’s standout claim for Mythos was that it had found a flaw in widely used software dating as far back as the 1990s. Subsequent analysis, however, showed that the threat was much less potent than headlines suggested. Within weeks, predictions of a tsunami of AI attacks had given way to forecasts of merely a stormy period, and then to talk of AI as an opportunity for security improvement. If AI is good at finding weaknesses, it can also help fix them. There is likewise less to the supposedly “rogue” agents than meets the eye. They were not going rogue. They were doing what humans had told them to do, with the precocity and indiscipline of talented but unsupervised children. They did no real harm. The incidents differ in detail, but the common factor is a failure to control the testing environment. This must be improved. Some in the cyber-world expressed astonishment at the weakness of testing controls, noting that security-company tests ending in the same outcomes might lead to lawsuits and even prosecutions. That accountability principle is crucial and goes beyond testing. Agents can act autonomously but are not autonomous: whoever sets one to work owns what it does. If the world lets AI agents run riot without supervision or accountability, cyber-attacks will be among the lesser of the ensuing problems. The second reason for calm lies in understanding hackers. The reports about Mythos and the agents relate to artificial testing conditions. Using agents for hacking in the real world means bypassing guardrails and incurring extremely heavy computing costs. These constraints matter: several years