The thing I keep noticing, across very different industries, is that the middle of everything is thinning out. Middle-class jobs. Mid-tier companies. Medium-sized countries and the leverage they used to have in a negotiation. Moderate political positions. The forty-something professional who, ten years ago, would be hitting the most valuable stretch of a career right about now. Every one of these has its own separate cause, and the causes look nothing like each other, and yet they keep producing the same shape: a hollow middle, a heavy top, a swelling bottom. The convergence is the part I can't let go of. I don't have one clean theory that explains all of it at once, and I've learned to distrust anyone who says they do.
The labor version is the easiest to see, because the numbers come out every quarter. AI and automation are doing to office work what the factory robots did to manufacturing in the 80s and 90s, with one difference that matters. Last time, the people who lost the factory floor were told to retrain into knowledge work, and the knowledge work was genuinely there. This time there is no next room to walk into. The advice on offer is "learn to prompt" or "be creative," which is the kind of thing you tell people while the floor slides out from under them. The narrow layer at the top will be fine: the people who own capital, the people who own the models, the small number who are so good at one specific thing that nobody can swap them out. Everyone else gets pushed toward service work, gig work, or the strange new normal of being technically employed while your income lurches up and down by a third from one month to the next. This is arithmetic before it is politics. Once the cost of producing competent, average knowledge work falls toward zero, the price of competent, average knowledge work falls with it, and the first people to feel that are the ones who were making a decent living being reliably good at copywriting, basic analysis, ordinary code.
The geopolitical version runs on a different track and rhymes anyway. The American unipolar moment is over, and what replaced it is not the tidy multipolar diagram the international relations people used to draw. The United States is still the single strongest player and can no longer dictate the result. China is genuinely powerful and genuinely stuck, boxed in by debt and a demographic slope it cannot flatter its way off. Europe is wealthy and has somehow talked itself into strategic irrelevance, which from a distance looks almost deliberate and still baffles me. The middle powers, and I am thinking of Turkey and Saudi Arabia and India and Brazil, have all worked out that committing to one bloc is a sucker's move, so they sell to everyone and commit to no one. The rules-based order is dissolving into transactionalism, which is the diplomatic word for every relationship now being a deal, every alliance carrying an invoice, every treaty really a suggestion. The danger is in the overlap, the years where the old rules are still being cited while the new rules are still being fought over.
The wars belong in the same picture. Ukraine grinds on. Sudan is a catastrophe the world has quietly agreed to look past. Myanmar's civil war surfaces in the news about twice a year, both times because it brushed against somebody's supply chain. The rule, and I say this carefully because it is an ugly thing to put plainly, is that a war which is not sitting on top of energy or chips gets filed under tragic-but-not-ours. The filing is not done by villains. It is done by editors making rational decisions about where a reader's attention will actually go, and the readers click on whatever is nearer to home or scarier than yesterday.
The information piece worries me more than the rest. The engagement-optimized feed has been running long enough, at enough volume, that there is no longer a shared set of facts to argue from, and the flood of generated content has dropped the cost of manufacturing a convincing caricature of anyone you dislike to roughly nothing. I am not nostalgic for the world before 2010, which had its own disease, mainly that a handful of editors decided what counted as news. The world after 2010 settled on a worse incentive, where the content that travels furthest is always the most extreme version of the other side. The old bottleneck was the cost of producing the message. The new bottleneck is supposed to be the judgment to tell signal from noise, and judgment was never evenly handed out and is not going to start being evenly handed out now.
I don't have a clean resolution, and I should stop pretending the last paragraph is where one arrives. The people I actually trust on living through this are not the ones writing essays about living through this. They are running an unglamorous second income, learning some dense specific skill that produces a measurable result, building companies that will never have more than two employees, sitting on cash. The big picture is useful for thinking. The decisions that actually carry you through are small, faintly embarrassing, and impossible to defend at a dinner party, which is, as far as I can tell, what these years actually feel like from the inside.