Some 2026 Predictions

I think we've made it past rock bottom?

We have nearly made it through 2025. (0/10, would not recommend). This year has been awful. It will take decades to repair all the damage.

I think 2026 is shaping up to be... better? Not good, mind you. But certainly less-bad. Here are my hunches about the year ahead:

Photo by Matt Walsh, via Unsplash: https://unsplash.com/photos/text-tVkdGtEe2C4

---

• By late spring/early summer, Trump’s governing coalition will be consumed by infighting.

This won’t lead to anything dramatic like impeachment proceedings or invoking the 25th Amendment. But one of the critical differences between Trump’s first and second administrations is his ability to quell any intra-party opposition. Basically no one within the party network has been capable to stand up to him this time around.¹ They have purged the ranks of the DOJ and the FBI and the military. The sole governing value that unites the Republican Party of 2025 is we agree with President Trump. The ability to maintain that party unity is one of his few relative strengths.

But, as Henry Farrell has argued, this sort of party unity is brittle. Absolute power can be a terrible weakness.

Here’s how it looks to me, right now: Trump’s approval rating currently hovers around 40%. The macroeconomic situation is dicey. Neither will improve in 2026. And, as we draw closer to election day, Republican leaders are going to panic and start blaming each other. Trump’s team will not deal well with dissension in the ranks, so things are likely to get out of hand.

That infighting will be a blessing, because the more time they spend fighting each other, the less bandwidth they have for attacking the rest of us.

This doesn’t mean we are out of the woods yet. Russ Vought and Stephen Miller are going to treat this year like a clearance sale on consolidating their fascist power grab. The Supreme Court majority will be even more brazen in trying to lock in election-proof permanent majority status for the Trump regime.

Still, assuming relatively normal elections in November, this could produce narrow Democratic majorities in both houses of Congress.

The internal fight among Democrats this year will be between the popularist “just say bland poll-tested shit and prepare to do nothing-at-all with your governing majority” wing and the twin left-populist/take-elite-corruption-seriously caucuses. The popularists will have a massive warchest supplied by wealthy donors, so their SuperPACs will dominate television advertising. But they’ll sound massively out-of-touch in ways that leave them falling flat everywhere else.

The popularist contingent will be represented by figures like David Shor and Matt Yglesias. They’ll sound a lot like Bret Stephens. I will spend mid-2026 engaged in far too many arguments with those guys.

---

• The AI bubble won’t burst quite yet, but it will become impossible to deny that we’re in the middle of a one.

I expect AI coding assistants will continue to make incremental improvements. I keep hearing that coders are blown away by the stuff. But that doesn’t mean generative AI is some magical general purpose technology. It might just be quite good for a few narrow sets of tasks.

AI slop is going to be everywhere in 2026, and everyone will fucking hate all of it. The cultural backlash will be in full swing next year. It has the makings of the biggest mass social rejection of a technology since Google Glass.

The AI finance bubble won’t quite burst, but only because the Trump administration (at the direction of David Sacks) props it up through a serious of absurd, obvious gimmicks. The government will supply public money to backstop huge Silicon Valley bets, recognizing that if the bubble were to pop before the November election, it would prompt a 2008-scale financial crisis. They’ll choose to prop up their VC buddies in the near-term rather than brace for the fallout.

But my hunch is that doesn’t save crypto. I think we slip into another crypto winter in 2026, as some of the big investors get scared about the AI bubble and reduce their exposure to other overinflated asset classes. There are still zero non-speculative, non-crime use cases for cryptocurrency. Once the big money get scared, all the valuations will go into freefall.

• Meanwhile, outside of electoral politics and AI…

-Billionaires buy 2-3 more major media outlets, consolidating their control of mass communication channels.

-Substack’s core newsletter product gets worse, as the company chases an endless series of growth hacks to keep their VC investors happy. Eventually they have another self-inflicted comms crisis that results in another wave of writers leaving for other platforms. I’ll be among them.

-We start a war with Venezuela or some other country Trump gets mad at.

-The mass mortality rates from preventable diseases ticks up, but it’s hard to conclusively prove the connection because the government doesn’t publish reliable data anymore.

-More carbon is pumped into the atmosphere than ever before, even though renewable energy keeps getting cheaper.

-A natural disaster strikes, and FEMA is woefully unprepared. The billionaire-owned major media outlets politely agree to barely cover it, and the algorithmic social media platforms suppress discussion as well.

-If there is a public health emergency, the government will just insist “no, there isn’t.”

---

I’ll end the predictions there, since this was meant to be an optimistic outlook and now I’m speculating on mass-death scenarios. (I am no fun at parties.)

Still, the bottom-line takeaway is that, if 2025 was defined by things falling apart faster than expected, my hunch is that 2026 will be defined by the bill coming due, and the Trump coalition fracturing as they are unable to pay it.

From there, the long road to recovery begins. It won’t be easy. But there is, I think, some light at the end of this tunnel.

1. Marjorie Taylor Greene stood up to him, so now she’s resigning from Congress. What an absurd year it has been, that such a sentence can be written as fact.