It has been a dark and stormy year, although not in as benign a sense as the much-mocked melodrama by Edward Bulwer-Litton from which this paraphrase of his opening line derives, nor as funny as the Peanuts cartoons that played on it. The year has blended dangerous geopolitical turmoil and extraordinary political destructiveness with a powerful surge of technological development that could presage either prosperity or calamity – or perhaps both.

As such, for this author the year’s principal themes are best evoked not by the purple prose of Victorian gothic fiction but by one famous quotation from an Ancient Greek historian and two famous book titles from past eras.

The Greek historian is Thucydides (who else?) and the quotation his poignant description of the might-is-right era of the Peloponnesian War, nearly 2,500 years ago. That was a time, as he wrote, when some felt that ‘the strong do what they can and the weak suffer what they must’.

This, certainly, is how the great powers of our time, Russia, China and, most shockingly, now the United States have been behaving. They believe that their strength and size entitles them to behave in ways smaller fry cannot. The rest of the world, they think, just has to put up with it.

Until 2025, it was generally believed that, for all its faults and occasional bullying behaviour, the United States had stood behind the international laws and norms that since the 1945 United Nations Charter have sought to protect the weak and constrain the strong. But in his second term in office President Donald Trump has shown no interest in laws or norms. Far from resisting such great-power behaviour, he has embraced it, making threats against his benign neighbour, Canada, casting greedy eyes on Greenland, an autonomous territory of his NATO ally, Denmark, and treating everyone, past friend or not, as suitable targets for tariffs and mafia-style shakedowns.

Dean Acheson served as a foreign policy adviser to President Harry Truman in 1945-47 and then as America’s Secretary of State in 1949-53, which was the period when such laws, norms and institutions were being built, under American leadership. The aim was to help avert any repeat of the devastating great-power rivalry of the first half of the 20th century as well as to constrain the potential destructiveness of the Cold War.

His memoir of his years at the State Department, Present at the Creation (1969), came to encapsulate the official American mindset of that period. It showed how even at a moment when America’s military and economic superiority over the rest of the globe was at its highest, that country sought to establish global leadership by creating laws, institutions and alliances rather than by using brute force. This year President Trump has set out with zeal and determination to destroy or disable many of those American-built institutions and alliances.

It is no exaggeration to paraphrase Acheson by saying that in 2025 we have been ‘Present at the Destruction’. This has been true most radically of the rules and norms that have governed international trade since the General Agreement on Tariffs and Trade was signed at America’s behest in 1947. But it also applies to the leading role that the United States had played in overseas aid since President John F. Kennedy founded the USAID agency in 1961 and to the destruction of the trust, sense of shared interest, and habits of consultation that lay behind America’s security alliances in Europe and Asia.

The second American withdrawal from the Paris climate agreement under a Trump presidency was not a surprise, but the Trump administration’s disregard of national and international laws of war, in its own actions as well as those of Israel and Russia, did count as one. This falls under the ‘might is right’ heading, though it has also reflected the dramatic expansion of presidential power that has taken place domestically in America under Trump, some of it ratified as constitutional by the US Supreme Court but some dependent on a compliant and intimidated Congress being under Republican majorities in both houses.

These two themes have made for grim viewing. That is especially true for those in actual or potential war zones who had previously depended upon American support and have now seen the United States throwing its own weight around in South America and Iran, and ceasing to act as an effective constraint on Russia in Ukraine or on Chinese bullying, thankfully short of war, in the Indo-Pacific.

Trump has tried to paint himself as a peacemaker. Yet two of the agreements he has claimed credit for (those between Cambodia and Thailand and the warring parties in the Democratic Republic of Congo) were crumbling within months or even days, and in two (India and Pakistan, and a clash between Egypt and Ethiopia over water) the parties dispute American involvement. Even the deal between Israel and Hamas looks fragile and is currently failing to make progress towards an enduring settlement.

Yet alongside that sense of gathering clouds and impending doom there has also emerged the mother of all technology booms. Like all such booms this has been founded on optimism. The race to develop and exploit Artificial Intelligence (AI) has driven economic growth in 2025, especially in the United States, chiefly through capital investment in computing power, and has led to spectacular stockmarket valuations, through speculation about how (and by whom) economic and social life promises to be transformed by the new but still little-understood technology.

That is why the other book that this year brought to mind was one from 1841, by a Scottish writer called Charles Mackay. His Extraordinary Popular Delusions and the Madness of Crowds mapped the way in which, over the centuries, innovations, especially ones cloaked in mystery, have led to financial manias and then crashes, whether of stockmarkets or banks or both. Since Mackay there have been many fine books about manias, for there have been many more booms and busts, but his really set the standard for all the rest, at least in the English language.

This is not to say that the fact that during 2025 an estimated 40 per cent of American economic growth can be attributed to spending by AI developers on data centres and other infrastructure necessarily means that AI represents a popular delusion. But, like the European and American railway booms of Mackay’s lifetime, or the South Sea Bubble a century earlier, or the ‘dotcom’ boom of internet and telecoms companies during the late 1990s, a vast amount of money has again been chasing unknown — and unknowable — sources of future profit.

When it is impossible to get hard facts on where the money is going to be made, it comes to seem rational just to follow the crowd and even eventually to share the crowd’s madness. In the 1720s, when speculation on the riches supposedly to be made, many thousands of miles away, by the South Sea Company hit fever pitch, even the greatest scientist of that time, Sir Isaac Newton, got caught up in the fervour and lost money when the stock crashed, though quite how much he lost remains a controversy among historians.

During 2025 there was plenty of evidence that AI can do remarkable things, but little sign of it boosting the productivity of large enough categories of users to make a difference to overall national income. At the same time there were manifold signs of major technology companies raising huge borrowings to finance their AI investment rather than funding it out of income, which means that if the market starts to crash their falls could be harder than most.

It is not surprising that worries about an ‘AI bubble’ became so widespread during the second half of the year. The technology is in that state summed up by the brilliant futurist Arthur C. Clarke as being ‘essentially indistinguishable from magic’. But also, to borrow another of Clarke’s famous sayings, it looks to be in the typical phase when its impact may well be being over-estimated in the short term while under-estimated in the long term.

During 2026 we may discover who has been the Isaac Newton of the AI bubble. But this doesn’t mean that the potential benefits of AI are a myth. In their early years, railways were over-estimated and over-built. But the railway built in Mackay’s time just beyond the bottom of what is now my Dublin garden is still running and has contributed greatly to productivity and national income over the ensuing 175 years.

It is too soon to judge how much this technological magic will do to counter the other bleaker trends in our world. If we want to be gloomy, we might compare today’s evidence of human genius to the way a great surge of technological advancement accompanied the depressed and deadly years of the 1930s and 1940s, from penicillin to jet engines to the transistor, albeit with the dropping of the atom bombs on 6 and 9 August 1945 representing the most awe-inspiring but also terrifying fruits of that genius.

Less apocalyptically, we could alternatively compare 2025 to the railway and oil booms that generated the huge wealth and political influence of the ‘robber barons’ in America at the end of the 19th century, for the AI boom has similarly turbocharged the concentration of wealth in the hands of today’s technology oligarchs. Just as that early 20th century concentration led to new antitrust laws and monopoly-busting politics, so might this one produce a beneficial backlash.

Finally, to look on the even brighter side, we might hope that with the boost that AI could bring to productivity and living standards could, at some point this decade or next, lift the developed economies out of the funk that has given rise to simplistic populist solutions from the likes of Trump, Farage, Le Pen and others. Against that we have to place the potential for AI to cause job losses and worsen inequalities of income.

However optimistic we may wish to be, we cannot hide from the fact that, much like during the 1930s, the raw use of power, both military and economic, came to the forefront of geopolitics this year, though that trend was scarcely new. Russia’s now nearly four-year-old attempt to re-conquer its former colony, Ukraine, nonetheless made little progress, with Russian forces expanding the Ukrainian territory under their control by less than half a per cent of the country’s total land mass at the cost of at least 100,000 Russian deaths and perhaps four times as many wounded, according to estimates published by The Economist. No major towns or cities have changed hands.

If it lasts beyond mid-2026, Russia’s war will have gone on for longer than did the First World War. However, President Vladimir Putin did make political progress by finding an American interlocutor in President Trump who is uninterested in Russian war crimes and disinclined to provide material support to Ukraine, unlike his predecessor, but was happy to rehabilitate Putin by inviting him to a summit on American soil in August and willing to treat as legitimate Putin’s maximalist demands as the price for peace.

As the year ended, there was no sign that Putin is willing to make the sort of compromises that a peace settlement short of victory will require, nor any sign that Ukraine’s president, Volodymyr Zelensky, is willing to capitulate. The European Union had struggled all year to cope with America’s change of stance under Trump, seeking constantly to persuade him at least to maintain minimal levels of intelligence and communications support for Ukraine and to convince him that Russia was not on an inevitable road to victory. The 19 December decision by the EU to provide Ukraine with €90 billion of financing, sufficient to keep it afloat and able to produce weapons for a further 1-2 years, should the war continue for that long, represented an important contribution to Ukraine’s continued ability to say no.

The hardest aspect to interpret of this might-is-right behaviour has been that of whether America under Trump wishes to compete with the other superpowers – as had previously been assumed, especially with regard to China – or whether it wishes in effect to divide the world between them into spheres of interest. Trump’s preference to seek an accommodation with Putin rather than to confront or constrain him could be interpreted as an effort to wean Russia away from the partnership with China that Putin and President Xi Jinping announced in February 2022, three weeks before the full invasion of Ukraine, and thus as part of the US-China contest. Or it can be interpreted as a belief that American power is incapable of significantly altering the behaviour of either Russia or China at an acceptable cost, making accommodations with both Putin and Xi look like the preferable course.

Certainly, 2025 must be counted as a good year for China, not because it prospered or achieved any important geopolitical aims but rather because of what didn’t happen. Its partner, Russia, did not collapse; America’s attempts to bully it with trade tariffs and controls on technology exports crumbled surprisingly quickly once China pushed back hard with its own controls over exports of critical minerals; and rather than confronting China over its ‘wolf warrior’ diplomacy and its military expansion in the South China Sea or the Himalayas, some of its Asian neighbours reluctantly sought friendlier relations with it once they saw that Trump was going to hit them with tariffs and extortionate demands.

This does not mean that China has expanded its network of alliances. It remains the case that the overwhelming preference of the world’s small and medium-sized powers is to preserve their autonomy and not to join the ‘camp’ of any great power, unless like Japan and South Korea they are already part of a long-running security alliance with the United States. However often the label may be used, there is no such coherent grouping as the so-called the ‘Global South’. But what the disparate and vast range of countries under that label do share is a policy of non-alignment.

During the 1950s, when the then ‘Non-Aligned Movement’ was formed as a grouping led by Yugoslavia and by former colonies such as Indonesia and India, the spirit that drove the movement was born out of weakness, a weakness that inspired a hope that by working together that weakness might at least be mitigated. Today’s non-alignment is born more often out of strength: the sense that thanks to successful economic development in this age of globalisation, many countries have more agency and more options.

Which is also in part why another feature of 2025 was a dog that didn’t bark: the dog of deglobalisation. Trump’s signature economic policy has been one of destroying the postwar system of open trade, commonly agreed rules and the principle of non-discrimination. His ‘Liberation Day’ on 2 April was in truth a day of destruction. But this American example was neither emulated by others nor retaliated against in the sort of escalating trade wars that were seen the last time the United States imposed high tariffs, in the 1930s.

The only country that engaged in serious retaliation against Trump’s trade measures was China, and its imposition of matching triple-digit tariffs and export controls on critical minerals led him to back down. The European Union chose not to retaliate, even though it has sufficient economic power to do so. And no other country nor trading bloc has chosen to follow America down the protectionist path.

The result is that while the system of rules and institutions has been seriously damaged, world trade has not been gravely harmed. Trade in services, which makes up nearly 30 per cent of global trade, was anyway not affected by tariffs and its growth is being fuelled by the internet and AI. Merchandise trade with the United States has begun to decline, although, following a host of bilateral negotiations, the effective average tariff imposed by America has settled at a lower level than looked possible on 2 April. Nonetheless, at an estimated 15 per cent, it is still the highest it has been for more than 80 years. Most importantly, trade has been diverted rather than destroyed, with supply chains and markets adjusting gradually to the new circumstances and middle-income countries benefiting as a result.

America is important, as an economy representing 25 per cent of global GDP, but its share of global trade is much smaller at 13-15 per cent, a share Trump’s tariffs are designed to reduce. What has been happening, therefore, has not been deglobalisation but rather de-Americanisation. The big question that has been frequently asked in the rest of the world, but not yet resolved, is whether de-Americanisation can now be extended to the rules, norms, institutions and even alliances that America built and has now chosen to disown, with those rules being maintained and even enhanced without America’s participation.

In November, Singapore hosted the inaugural meeting of a new body, the Future of Investment and Trade Partnership, assembling 16 small and medium-sized countries to start talking about preserving the global trading system. That is a start, but has yet to involve the biggest trading blocs. The EU talks about negotiating new trade agreements with other blocs, such as the Comprehensive and Progressive Trans-Pacific Partnership that groups together 11 Asia-Pacific economies as well as the UK. Yet after 25 years of negotiation, the EU still failed in December to sign off on a new deal with Mercosur in South America, deferring the issue until 2026.

As the Trump administration has made its nature and intentions clearer, so attention has begun to shift from short-term damage-limitation to the creation of arrangements that can endure for the longer term. Yet honest governments that have in the past enjoyed close relationships, even of dependency, with the United States, find themselves caught in a dilemma: while no one has ever lost money by betting on Donald Trump’s bad faith and malign intentions, the duration of what might be termed the Trump Supremacy is still being assumed by policy makers to be a great deal shorter than the planning horizon required for building true autonomy in defence, trade or anything else. America feels like an ex-partner, one that can no longer be trusted and is often coercive and abusive, but still one that has to be lived with.

Moreover, it cannot be said that the governments of those former partners in the now defunct West, such as the UK, France, Germany, Japan or others, feel strong and confident. All suffer from economic frailties, in Europe’s case considerably aggravated by the Ukraine war, and many have seen domestic populist parties, generally nationalist and anti-immigrant, rise in the opinion polls. Strategic autonomy, to use the phrase popularised by President Emmanuel Macron, looks more desirable than ever but also far out of reach.

Hope may not be a strategy but it is a common human sentiment. The mid-term Congressional elections of November 2026 will reveal whether America’s democracy and its constitutional guardrails remain intact or whether the ‘America First’ period is destined to last for much longer. Never in my professional lifetime have mid-term American elections looked more consequential, both for the United States and for the world.