Year End Review of Things Pt 2
Text and Predictions
Best of the From the New World Newsletter:
I don’t think I got the ball rolling in articulating my political theory in better detail until the end of the year. Consequently I think the post from last week was actually the best one. I’m also very excited about the post next week, on Balaji Srinivasan’s The Network State. In general, I’ve neglected the more general theory in favor of more specific writing on events, news stories, or other posts, some of which have been good.
Stupidity beats intelligence. I just explained why I think this is my best article of the year.
Sadly, Porn: A Future that Forgot History. This book has very real practical impacts on political strategy. I’m not joking. I wish more people would realize this. Relevant quote:
People don’t care whether culture war stories are made up or not. They are fiat stories, made valuable solely by their usefulness. Here is a good example:
I don’t know if this post is made up. It in fact doesn’t matter at all if this post is made up. The random woman and her dog doesn’t affect your life in the slightest. And so it goes with most culture war anecdotes, from George Floyd to Libs of Tiktok. The reaction to the story IS the story. The ability to chant a few syllables and rally a partisan army is the story. Without that power, George Floyd and the random LoTK teacher would matter as much to you as the 100th car crash victim this year. Know his or her name? I didn’t think so.
This is somewhat postmodern, but it’s important to understand there is an underlying reality, even if it is less powerful in affecting media or politics in the short term. Even if the general population cannot believe in history or statistics, those exist in reality. Inflation, geopolitics, supply chains, and technology are real things with real impact, whether media covers them or not. It is the duty of people who understand this to also understand the first point. It is our duty to align truth and media power. It is their duty because they are, tautologically, the only people who can.
And of course this list would not be complete without my ongoing series of reporting on OpenAI language models. Part 1. Part 2.
Books and Articles
You can take this as an extended reading list of books.
Unsurprisingly the substacks I read the most are Richard Hanania, Rob Henderson, and Zvi Mowshowitz.
This by The Free Press seems to me like the most important of the Twitter Files.
The Network State for reasons I will explain in a few days in part one of my review.
I have a 100% completion rate for Alex Nowrasteh’s substack, mostly because it has only two actual articles. I finally set up recommendations, which you can see here. Right now, it’s the substacks of everyone who has appeared on the From the New World podcast. I may add more in the future
Sadly, Porn if you can bear the writing style. I already explained why. Here are some reviews if you can’t bear the writing style:
Predictions about the World:
The FDA, CDC and other Public Health Organizations will continue to actively harm the public health by banning tests, treatments and vaccines. In the next 5 years, there will be some minor epidemic (similar in scale to H1N1, monkeypox, or flu/RSV), and the FDA will behave very similarly to monkeypox, banning initial tests and treatments (75% that there will be a minor epidemic, 95% that they FDA will worsen the public health if such an epidemic occurs). Contra Tyler, bureaucracies will not learn unless you destroy and recreate them.
Trump wins the Republican primary (66%) and loses the general (90% given Trump winning the primary). People really don’t change that much over time, past behavior is the best predictor of future behavior, etc. Do I really have to add anything else? A lot of my “Trump doesn’t win the Republican primary” percent is also extraneous events, like if he is barred from running, or if he drops out for some reason.
I also think people overrate the probability of DeSantis running at all. I usually see people refer to DeSantis running as if it's taken for granted, but I think it’s closer to 75%. Still likely, but far from Guaranteed. People raise the point that DeSantis is term-limited out of being Governor of Florida again, but that still allows him a reasonable timeline of serving out his term and then running for President in 2028.
China’s semiconductor research industry will not produce chips with superior benchmarks than the United States’ in the next 5 years (95%) or next 10 years (90%). This is a pretty uncontroversial take but I will use it to comment on the fact that Chinese semiconductor subsidies are in fact terrible industrial policy and probably made the Chinese industry irreparably fake. My father analogized it to the Western crypto industry in August and I think the sentiment of most people in or adjacent to the Chinese semiconductor industry is similar. Maybe betting on a major bankruptcy or bailout might be preferable, but this is difficult since the existing government support makes those companies actually profitable and it can last indefinitely.
Major tech layoffs continue throughout 2023. If I were to specify, at least a 10% layoff of white collar staff in a MANGA company in Q3 or Q4 2023 (66%). I think that once executives and investors start paying attention to how they can course correct, they will continue to repeatedly find large groups of workers unnecessary. If I were to bet on one company specifically, it would be Meta/Facebook (33%, or half of the total probability).
Predictions about From the New World:
At least one FTNW guest goes on Joe Rogan in 2023 (80%). Leading contenders are Michael Shermer, Rob Henderson, Eric Kaufmann, Curtis Yarvin, but honestly these things are so high variance that I couldn’t really give a precise answer for any one guest.
This is what my free subscriber graph looks like so far
So if I were to predict free subscriber growth, I would say around 20x (50%). This is notably 2x than my twitter, prediction since I think I will be investing much more time into substack and not that much more time into twitter. I’ll also keep the absolute numbers private for now, for no reason in particular.
I’ll get to 50,000 (10x) followers on Twitter (50%). I think 50000 is a reasonable over/under. I think it’s pretty high variance depending on what stories I break and appearances I have on other podcasts.
I’ll break a major story about something political OpenAI does in 2023 (90%). My (limited) experience is that no one else breaks these things, so this is mostly a bet on whether OpenAI does something political in 2023. Of course “major” is subjective but basically on the same scale as my first article.
Openly soliciting bets about FTNW. I may make a manifold market on this in the future.
I don't think modeling your growth as exponential is the right assumption. Based on what I've seen (I think Noah Smith had a post outlining his subscriber growth?), linear growth is closer to correct.
Though of course I will be rooting for exponential growth!