Trump is also using every possible opportunity to show them that he will pardon them for any crime they commit for him or for themselves while serving him. Time is not on the side of the Democrats here. The regime must not be allowed to solidify and consolidate, it must be disrupted to the maximum.

I think this is an essential analysis in other to defeat the new fascists. The cultural and ideological parts of fascism are harmful but that is neither their strength nor their weakness - and to focus on it is a mistake. We must attack the money. I hope to have a longer substack article on it soon.

Richard Jevans: "one of the major and too often neglected factors holding the top ranks of the Nazi Party together was corruption: the receipt of massive gifts, huge increases in income, property and so on, that cemented the patronage relationship they had with Hitler in a relationship of clientage”

Paul Corner on Mussolini: "‘In a regime that rested heavily on a whole network of personal contacts ... corruption represented a kind of glue, keeping the network together… this discretionary nature of decision-making that rendered people so vulnerable to what was, in effect, blackmail…’

Götz Aly: "The Nazi leadership did not transform the majority of Germans into ideological fanatics ... Instead it succeeded in making them well-fed parasites. Vast numbers of Germans fell prey to the euphoria of a gold rush, certain that the future would be a time of unbridled prosperity."

My conclusion is that it is a modern version of something that has been around for as long as have historical record: Pillage and plunder. The nazis and the fascists both immediately maxed all ways to rack up national debt, plundered their own people, then looted other countries. Some quotes:

Fascism should not be understood as an ideology. Any particular instance of fascism is just an arbitrary confabulation of its leaders, some crackpot theory they have decided to latch onto, that is not necessarily shared with any other instance. Fascism, I realized, is above all an economic movement.

Reading Götz Aly and Adam Tooze on the economics of Nazi Germany in the evening while hearing about the kleptocratic smash and grab of 12% of an entire country's GDP in the morning finally made me connect the dots. These are not different movements, they are the same thing.

I usually read the most while traveling. For me, sitting in an airport terminal, in a hotel lobby, or on a train is perfect for concentration. The journey took me as usual from country to country in rapid succession. Eventually I came to Moldova, where I was told of the great bank robbery of 2014.

Savages, who puts ice in their coffee?

It is also nice to see a discussion of how we should define a more general interface between academia and government, where government can temporarily draw in academics while they keep their jobs. We need more of this.

Our collective information space needs to be seen as an organism - it needs active repair processes. The tech overlords cannot supply this - there is no profit in it. Only academia can provide a solution to this - and save AI from itself. And save all of us from living in a disinformation hell.

On top of that we have all the the active attempts to manipulate and poison such algorithms. Before, we had the search engine optimizers making havoc in the search space, but now that they can use AI to manipulate AI, this poisoning process goes into overdrive.

Worse, as it starts to feed on itself, as its outputs become its own inputs, it becomes poisoned. The discrepancies and the inaccuracies, they all accumulate in the lack of active repair. The information universe tends towards entropy and chaos.

The "Attention Is All You Need" paper is the fundamental basis for all the current AI craze. But in the end it just does for smaller pieces of information the same as PageRank did for pages - it blindly smashes them together using other people's organizing work. It is a limited by its availability.

Unlike their competitor Yahoo, they did not attempt to organize information themselves, they just combined everyone else's organization of information on the internet into one big engine. This was the genius of their PageRank algorithm. Other algorithms followed, and now large language model paper.

The main problem facing the veracity of AI these days is not a lack of information, it is the lack of organization of information. Doing this by algorithm was Google's mission - they got incredibly rich by leveraging other people's work to do this organization.

The current idea of papers being reviewed only once and then published once and for all is silly in the information age. Much of what is wrong about a paper is found later, and then you will have to find about that in other papers or online articles.