Age of AInxiety
As technology hammers children's mental health, AI risks destroying the economy as we know it
Towering Columns
In The Times, Juliet Samuel says tech elites are removing their children from smartphones while social media companies damage ordinary children’s brains.
Unsurprisingly, given these methods monetise eyeballs in the way social media does, they are very “engaging”, which is to say they mesmerise the developing brain with stunning power. But as [Jonathan] Haidt points out, they also do two other things: they damage the ability to concentrate on anything that doesn’t deploy addictive methods; and, as a result, they have generated a sustained fall in objective measures of achievement. Pisa scores, basic numeracy and literacy — the declines are greatest for the most deprived. This precedes Covid, dating instead from the start of the smartphone era. In other words, most of what badges itself as ed tech “develops” the brain only in the way that sugar “develops” the body, not by building capacity and strength but by stimulating and feeding mental cravings.
This doesn’t even take into account the “device controls” runaround, the ridiculous ratchet of spending and reassurance by school IT departments that claim to have invented infallible systems to contain distractible students. The kids are meanwhile busy filling up Quora, TikTok and wherever else with tips on how to beat the controls using VPNs and play games from which you can “quickly alt-tab away when the teacher teleports behind you”. (Some good ones, this “experienced gamer” recommends, are Krunker, Slither and Agar, which also sounds like a law firm for rotted distraction-addled brains).
Among the super-elites of Silicon Valley, by contrast, there is a growing fashion for keeping children off social media, strictly limiting device time and paying thousands for screen-free schooling. Ed tech, you see, is only for the techno-bamboozled chumps who can’t see the problem with dropping their darlings off for a full-day of screen-based zombification followed by an hour or two of app quizzes, homework tapping and message frenzies before bedtime. Looking around London primary schools a few years ago, I found the poshest were often the proudest of their iPad stacks, individually labelled for each six-year-old. Even for parents alert to the problem, avoiding this stuff has become impossible. The backlash is on its way, but ever so slowly. On top of getting smartphones out of schools, the campaign group SafeScreens is now demanding “a statutory right for parents to opt out of digital homework and digital-first learning”. Brigades of parents at their wits’ end are on the march. They can see with absolute clarity that the tech overdose isn’t just making the kids irritable and anxious; it’s making them stupid and unemployable to boot.
Also in The Times, Robert Colvile says activist judges are effectively legislating over policy areas from immigration to planning and equalities laws.
Indeed, asylum and migration is widely accepted to be one of the most left-wing parts of an increasingly left-wing profession. In surveys by The Lawyer, the magazine’s readers voted overwhelmingly for the Lib Dems in 2019, and even more overwhelmingly for Labour in 2024. But immigration isn’t the only area where the idea of the judiciary as impartial interpreters of the rules set by parliament seems increasingly antiquated. A few months ago, every living lord chief justice co-authored a report on prisons policy, urging more leniency and lamenting the fact that idiot politicians, egged on by the media, kept pushing for tougher sentencing.
Elsewhere, judges have repeatedly blocked oil and gas projects because of the carbon they would generate, despite a long-established precedent that such “indirect emissions” should fall under national climate policy. The nutrient neutrality scandal has seen tens of thousands of houses blocked by a disastrously strict interpretation of habitat regulations. And the government has yet to explain why we are legally obliged to give up the Chagos islands, let alone hand billions to Mauritius for the privilege. Not least since, when we signed up to the International Court of Justice, we explicitly excluded such disputes from its remit, precisely to head off this kind of grift…
…[L]awyers have constructed an incredibly elaborate apparatus to determine what constitutes work of equal value. In the recent verdict against Asda (on similar grounds to Next), the judges spent more than 300 pages going through every task workers were asked to perform, to determine the level of training required, the precise amount of physical and mental exertion, and so on. The same legal constraints have seen the NHS grade all staff according to an impossibly complicated formula that creates arbitrary equivalence between completely different roles. What is missing from such systems, or at least drastically underweighted, is the most important thing of all: what the market will actually pay. For example, the evidence in the Next case was very clear that there is more demand for warehouse workers than retail staff, that pay for such roles is higher, and that the firm had begged staff in its stores to shift over. Likewise, there are pretty obvious reasons why Birmingham might have needed to incentivise bin men more than those doing a 9 to 5. Rather than reflecting reality, the equal pay regime has become a charter for trade unions to milk employers. And I’ve explained it at such length because Labour is hugely expanding its scope.
On Commonplace, Michael Lind argues that free trade absolutism is an unholy alliance of ideology and corporate self-interest.
Libertarian ideologues are another group who bring a degree of emotionalism to technical debates about trade that is far out of proportion to the stakes involved. Although they are considered to be on the political Right today, in the 19th century their predecessors were called liberals or radicals. They opposed not only protectionism but also empire, colonialism, military spending, guilds and labor unions, and public welfare programs. Liberals, anarchists, and Marxists shared similar Enlightenment assumptions about natural human sociability and rationality. They agreed that the goal was a post-national global society of radically-emancipated individuals, although they disagreed about how to get there.
Why do today’s libertarians tend to devote nearly all of their time and energy attacking tariffs and other trade regulations rather than fighting on other fronts in the war against government? The answer is that self-interested corporate donors and wealthy individuals manifestly spend more funding libertarian thinks and pundits to promote free trade rather than funding other libertarian causes like drug legalization. The alliance of convenience and (sometimes) conviction between economic interest groups that favor trade liberalization and devout free-market radicals first coalesced during the debate about the Corn Laws in Britain in the mid-nineteenth century (“corn” is the British term for what Americans call “wheat”). The industrial capitalists of Manchester and other British manufacturing centers—the “Manchester liberals”—sought to maximize their profits while minimizing the wages they paid their workers…
…However, the factory workers and their families needed to eat. By repealing the protectionist tariffs that protected British wheat growers in order to import cheaper wheat from foreign countries, the British government could make bread more affordable for the proles, even if their wages did not go up. Although most 19th century liberals opposed trade unions, calling for imports of cheap foreign grain allowed them to pose as friends of the working classes, in the same way that today’s free traders praise cheap Chinese imports for effectively increasing the disposable incomes of working-class Americans. In addition to helping to keep the cost of wages low, free trade also promised to increase the supply and lower the cost of cotton and wool for textile manufacturing, which was the most important industry in Britain at the time.
Also in The Times, Simon French says UK energy policy is a major deterrent to foreign investors.
Investors fully recognise the global imperative to decarbonise but in reducing domestic supply of oil and gas the UK is, in fact, adding to carbon in its energy mix. It does this by increasing its reliance on dirtier foreign imports. Approximately a third of the UK’s gas demand comes from imported liquefied natural gas from the US and Qatar. At 79kg of carbon per barrel of oil equivalent (BOE), this has almost four times the carbon intensity of UK production at just 21kg/BOE. The processes of liquefaction and transportation across the seas are both carbon-intensive. A straight replacement of imported LNG by domestic supply would diminish the UK’s total carbon footprint and reduce the offshoring of UK emissions.
The second irrationality is the perverse fiscal implications of limiting North Sea exploration at a time of tight public finances. Reduced domestic production raises decommissioning charges for taxpayers, estimated at £24 billion over the next decade, and reduces Treasury tax revenues. The Laffer curve, where a hypothetical cut in the tax rate generates more in tax revenue, is often cited in UK economic discourse with little empirical support, but at 78 per cent — the marginal tax rate on North Sea energy profits — we are in clear Laffer territory. At a time when the government is facing fiendishly difficult decisions on public spending and tax, leaving tax revenue on the table is an odd approach, to put it mildly.
Despite the UK government recently having opened a consultation on the future of the energy profits levy beyond 2030, the chancellor’s leadership of the growth mission and sound public finances will have a hollow ring if that tax persists in its current guise beyond the autumn budget. There is a final impact on the public finances that is most maddening of all. The UK public balance sheet will not have been short of requests in the current spending review. From schools to roads, from prisons to hospitals, there are arenas where public capital is essential to plug clear market failures and underinvestment, but energy is not one of these failures. The private sector is highly proficient at funding all parts of the energy market from generation and through transmission to storage. The very real risk is that the spending review spreads scarce capital budgets too thinly. The funding for Great British Energy looks like an unnecessary fiscal folly.
In Foreign Affairs, Michael Pettis says Trump’s tariffs are a reaction to global trade imbalances created by China.
But whatever happens in the near term, this much is clear: Trump’s policies reflect a transformation of the global trade and capital regime that had already started. One way or another, a dramatic change of some kind was necessary to address imbalances in the global economy that have been decades in the making. Current trade tensions are the result of a disconnect between the needs of individual economies and the needs of the global system. Although the global system benefits from rising wages, which push up demand for producers everywhere, tensions arise when individual countries can grow more quickly by boosting their manufacturing sectors at the expense of wage growth—for example, by directly and indirectly suppressing growth in household income relative to growth in worker productivity. The result is a global trading system in which, to their collective detriment, countries compete by keeping wages down…
…in a highly globalized world where some states are more successful than others at suppressing labour costs, the result is an asymmetry in the demand for and supply of goods. Because businesses do not have to make products in the same places where they sell them, local labour costs become crucial to the competitiveness of manufacturers. Businesses that shift production to countries where labour costs are lower relative to workers’ productivity can produce goods more cheaply, making their products more attractive globally. In any given state, wage suppression puts downward pressure on domestic consumption while subsidizing domestic production. This results in a rising gap between production and consumption which, if it remains within the economy, must be balanced by raising domestic investment (which can further exacerbate the gap between production and consumption). Otherwise, the gap invariably reverses, either via raising wages or by cutting back on production. But in a globalized economy, there is another option: running a trade surplus. This allows the country to export the cost of the gap between consumption and production to trade partners. This is why, in 1937, the economist Joan Robinson referred to the trade surpluses that resulted from suppressed domestic demand as the consequences of “beggar-my-neighbour” policies…
…If globalization is to thrive, the world must revert to a kind of globalization where countries export in order to import and where a country’s production, consumption, and investment imbalances are resolved domestically—not foisted on to trade partners. The world requires, in other words, a new global trade regime where countries agree to restrain their domestic imbalances and match domestic demand with domestic supply. Only then will states no longer be forced to absorb one another’s internal imbalances. The best way to achieve this kind of globalization is to create a new customs union, along the lines of what Keynes proposed at Bretton Woods. States that join would agree to keep trade between them broadly balanced, with penalties for members that fail. But they would also erect trade barriers against countries that don’t participate in order to protect themselves from imbalances outside the customs union. Trade would not be expected to balance bilaterally, of course, but rather across all trade partners. Its members would have to commit to managing their economies in ways that would not externalize the costs of their own domestic policies. In that system, every country could choose its own preferred development path, yet it could not do so in ways that inflict the costs of domestic imbalances on trade partners. (Smaller, less developed economies might receive some limited exemptions from the union’s rules.)
In Jacobin, Ha-Joon Chang says wealthy economies like Switzerland and Sweden have pursued successful pro-manufacturing policies to restore their industrial base.
You ask whether there is a precedent of the countries regaining their manufacturing sector. And actually Sweden and Finland have just done that. For a variety of reasons in the early 1990s, these two countries experienced quite a significant decline in the share of manufacturing in their economies. In Finland, manufacturing accounted for 24 percent of GDP. By 1991, it had declined to 17. In Sweden, manufacturing as a share of GDP declined from 21 to 16 percent during the same period. But by the early 2000, Finland brought its manufacturing share of GDP back up to 24 percent, and Sweden raised its manufacturing share of GDP to 20 percent.
The same trend can be observed in Singapore. Singapore experienced quite a significant decline in manufacturing in the mid-1980s, from 27 percent to 20 percent. But by the mid-2000s, it had recovered back to 27 percent. By the way, Singapore, despite what people think, is one of the most industrialized countries in the world: in terms of per capita manufacturing output, it ranks in the top five globally. There’s an interesting myth about it being a service economy.
The most industrialized country in the world is Switzerland. You think that the Swiss are dealing in the black money from Third World dictators and selling cow bells and cuckoo clocks to American and Japanese tourists. Actually, it is literally the most industrialized country in the world, if you count in terms of manufacturing output per person. These countries have managed to revive their manufacturing industry, and since then they have declined a bit. But the lesson here is that these countries could do that only because they had a deliberate policy to revive manufacturing. What Donald Trump is trying to do is wishful thinking. Countries that have successfully increased their manufacturing output have deliberate policies to support manufacturing. In the Swedish and Finnish case, it also extended to retraining the workers made redundant because of the decline in traditional manufacturing sectors and then turning them into workers for new industries.
Wonky Thinking
Luke Drago and Rudolf Lane published The Intelligence Curse. The authors examine the existential risks posed by AI and make recommendations on averting catastrophes, diffusing ownership and democratising institutions in response to the coming “intelligence age”.
We will soon live in the intelligence age. What you do with that information will determine your place in history.
The imminent arrival of AGI has pushed many to try to seize the levers of power as quickly as possible, leaping towards projects that, if successful, would comprehensively automate all work. There is a trillion-dollar arms race to see who can achieve such a capability first, with trillions more in gains to be won.
Yes, that means you’ll lose your job. But it goes beyond that: this will remove the need for regular people in our economy. Powerful actors—like states and companies—no longer have an incentive to care about regular people. We call this the intelligence curse.
If we do nothing, the intelligence curse will work like this:
Powerful AI will push automation through existing organizations, starting from the bottom and moving to the top.
AI will obsolete even outlier human talent. Social mobility will stop, ending the social dynamism and progress that it drives.
Non-human factors of production, like capital, resources, and control over AI, will become overwhelmingly more important than humans.
This will usher in incentives for powerful actors around the world that break the modern social contract.
This could result in the gradual—or sudden—disempowerment of the vast majority of humanity.
But this prophecy is not yet fulfilled; we reject the view that this path is inevitable. We see a different future on the horizon, but it will require a deliberate and concerted effort to achieve it.
We aim to change the incentives driving the intelligence curse, maintaining human economic relevance and strengthening our democratic institutions to withstand what will likely be the greatest societal disruption in history.
To break the intelligence curse, we should chart a different path on the tech tree, building technology that lets us:
Avert AI catastrophes by hardening the world against them, both because it is good in itself and because it removes the security threats that drive calls for centralization.
Diffuse AI, to get it in the hands of regular people. In the short-term, build AI that augments human capabilities. In the long-term, align AI directly to individual users and give everyone control in the AI economy.
Democratize institutions, making them more anchored to the needs of humans even as they are buffeted by the changing incentive landscape and fast-moving events of the AGI transition.
In this series of essays, we examine the incoming crisis of human irrelevance and provide a map towards a future where people remain the masters of their destiny.
OpenAI's goal is to build artificial general intelligence (AGI), which they define as "a highly autonomous system that outperforms humans at most economically valuable work". The other AI labs—including Anthropic, Meta, DeepSeek, and Google—have made similar claims. The CEOs of these companies think they might achieve it in just a few years, some saying as early as 2026.
While the exact timelines are still in doubt, there is a very real chance AGI arrives in the next few years.
Consider the trendlines. In 2019, state-of-the-art AI models couldn't write a coherent paragraph; by 2023 they were doing as well as the average candidate in human bar exams.1 In 2023, the top AI models were resolving 4.4% of a set of real-world example coding problems; by the beginning of 2025 they were resolving 49%. Other AI systems recently scored above the 99.5% percentile of expert humans on competitive programming problems. On multiple-choice science questions selected to be hard for everyone but PhDs in that specific field to answer, AIs improved from barely better than random guessing in mid-2023 to better than the human experts by the end of 2024. General computer-use capabilities are lagging behind pure-text skills, but have gone from near-zero capability to closing half the gap with humans within the last year.2 Researchers have even demonstrated a new Moore's law: the length of tasks AI can complete is doubling every seven months…
On his Substack, Neil O’Brien says universities have become a vehicle for students from low-income countries to remain in the UK long-term.
Students from overseas have very different effects on different bits of our higher education system. The system is now characterised by complex financial flows between tuition fees for domestic students, tuition fees for overseas students (higher), and the different types of research funding.
Anyone attempting reform of the system will have to tread carefully: it resembles one of those complicated-looking bombs in movies where the hero has to decide which colour wire to snip.
For a long time there has been a discussion about “taking students out” of net migration figures. This might seem odd to an outsider: after all, students who come to the UK to study and then leave take themselves out of the net migration figures.
Of course the real issue is that increasingly, many don’t leave, and instead move onto a work visa; or marry someone here, or claim asylum - or just disappear. That’s why over the three years to June 2024 students from outside the EU added 632,000 to net migration, and their dependants a further 225,000 - so that’s 285,000 extra people a year overall, a city-sized number each year. Around 60% of dependants were adults.
The reintroduction of the right to work after studying has (obviously) brought more students to the UK for whom work and settlement are important motives as much as the initial period of study. That plus other changes like the new post-2021 system and the care visa have also tilted the mix towards developing countries, from which staying-on rates are higher. As the ONS has noted:
More non-EU+ students and their dependants have been staying longer, and nearly 1 in 2 transitioned to a different visa type after three years from YE June 2021; an increase from 1 in 10 after three years from YE June 2019.
Two things are multiplying each other: the balance has shifted towards students from poorer countries who are more likely to stay, and the staying-on rate from those countries has also increased, with a large majority now shifting to other visas to allow them to stay on (plus some more on top who don’t get a visa, but don’t leave)…
Quick Links
The UK’s growth forecast was downgraded from 1.6% to 1.1%.
Borrowing is due to exceed forecasts by £15 billion.
The IMF recognised for the first time that surging immigration may push down UK wages.
Leaked messages from employees of social media companies revealed knowledge of harm to children’s mental health.
1 in 6 foreign-born workers who arrived in the UK within the past 5 years are on a zero-hours, variable-hours or temporary contract, compared to 1 in 10 UK workers.
The former Lord Chief Justice, Lord Burnett, said the ECHR has veered into determining political decisions.
Spending on disability benefits is due to soar as claimants rise to 4.1 million by 2030.
It takes on average 8.5 years to get planning approval for a development project in Britain.
British tech startup Isembard received £9 million to start reshoring critical industries.
Chinese net exports of manufactured goods grew more than 25-fold over the last two decades.