Eight predictions for 2024

Read Max is never wrong

Greetings from Read Max HQ, where we are slowly moving past the trauma of 10 straight days without professional childcare and greeting the world with focus and tranquility. To kick off 2024, Read Max’s crack Forecasting and Special Augury Department has developed eight extremely scientific predictions for the coming year. A word of caution: Prediction is powerful sorcery, and these prophecies are likely to affect the world in difficult to foresee ways. However, Read Max absolutely refuses to be held accountable for them coming true, or not coming true, and plans to immediately forget all of the predictions, or cite them constantly, depending on which is more convenient for our purposes.

The chatbot bubble will burst

I’m not really sure how to make this a falsifiable prediction, which is just as well because it’s probably wrong. But nevertheless let me use this platform to express some pessimism about the future of the chatbot--especially the mostly free-to-use chatbot--as the main consumer interface for large language models.

Chatbots (like ChatGPT and Bing AI and Replika) have been incredible marketing devices for large language models, for the obvious reason that the entire user experience (the chat format, the design of the app, the character “ChatGPT”) makes you feel like you’re talking to a guy who lives inside computer (up to a point), which is what a century of science fiction has conditioned us to understand to be “artificial intelligence.” (For more on “the chatbot” and how it should be understood I highly recommend this excellent piece by Colin Fraser.)

But while the chatbot is impressive and sometimes spooky, I’m not sure it’s particularly useful for most people. Large language models overall quite clearly have uses of all kinds--autocomplete, summarization, translation, data-wrangling, shitposting--but is “convincing a character inside my computer to do something for me” really the best way to access those uses? ChatGPT and its peers can be cajoled into acting like useful products for some people, but that’s precisely the problem: Computers are annoying enough as it is; I (and I suspect most other people) don’t want to have to sweet talk (or threaten) them in order to get them to do the things I want well, especially when fairly often the result is that they’ll do the thing wrong.

And these chatbots are expensive to run (and, depending on the outcome of the New York Times lawsuit against OpenAI, may get even more expensive to train). As the novelty wears off and the publicity benefits recede (or even reverse), I can’t imagine comapnies are getting much juice get out of making them close to free for most people to use.

I don’t think chatbots are going to go away--they’re such a convenient and attractive synecdoche for “A.I.” that journalists and critics will continue to reference them, and at this point there’s such an entrenched ecosystem of X and YouTube hustlers encouraging ChatGPT use and prompt engineering that they’ll continue to appear as objects of content on social networks. But I don’t think this is a Facebook-in-2005 situation where we’re in the early years of exponential user growth. It seems a lot more likely to me that most people will end up “using” LLMs as the power behind relatively limited and straightforward features (“summarize,” “explain,” etc.) built into other apps.

Internet atheist culture will have a revival

Internet atheism originally went dormant because it became clear that most internet atheists were either (1) really annoying or (2) more “deeply Islamophobic” than really committed to “enlightment values” per se, but in the absence of internet atheists and their earnest commitment to arguing with people to the point of madness, huge populations of extremely annoying zoomer trad-Christians and “mystical” post-rationalists¹ have been allowed to flourish on the internet. The mass adulation of the pony-tailed guy who refused to debate the annoying Christian TikToker is a reflection of a latent desire to reintroduce a natural predator to counteract the unfortunate proliferation of obscurantism, mysticism, and religiosity online. (It’s also the most prominent example of a simmering but misplaced nostalgia for forum culture of the 2000s that I think is likely to get bigger this year.)

At least one of the big e/acc influencers on Twitter is going to have a full-on meltdown and lock his account

These guys all got too Twitter-famous too fast. This arc never ends well.

Biden will win the electoral college but not the popular vote

There are many reasons to believe that this is what will happen: Biden is doing terribly in national polls, but the demographic (and geographic) concentration of his losses is uneven. Just going off of recent state polls, it’s easy to imagine a scenario where he wins Wisconsin, Pennsylvania, and Michigan while still losing the national popular vote by a few hundred thousand. And while polls will probably shift as the economy improves and the election gets closer, Biden seems uninterested in doing anything that might improve his standing among the voters who disapprove of him.

But the most important reason to believe this will happen is that it’s the stupidest possible outcome out of any of them, which is the only necessary heuristic for prediction elections in the 2020s.

Twitter ads will get worse

I know this seems impossible, but that’s what they always say about Elon Musk’s companies.

Man City will win the premier league

De Bruyne back, Haaland back, Pep will figure out the midfield. “Stats guys” will tell you different but it’s really obvious to anyone watching that Liverpool are too goofy to pull it off and Arsenal just do not have the juice, the whole club stinks of loserdom. Nothing is more inevitable than City reeling off one of those Terminator streaks from here on out.

I’m going to recommend at least one (1) Alex Proyas movie in this newsletter, and not the one you’d expect (Dark City)

It’s either going to be the one where Nic Cage uncovers a new-age prophecy about the apocalypse or the one where Jaime Lannister plays the Egyptian god Horus.

1. For more on this, check out the panel I was on, embedded in this post.