This is standard today, but how recent is it? It looks like the industrial age to me.

How much of institutions is about solving akrasia and how much is about breaking the ability to act autonomously?

We get the word akrasia from Plato, but was he really talking about the same thing?

There is always the question of whether to study things bottom-up or top-down. These are bottom-up studies of what to do if you have a single infected patient. If you had an individual infected with a novel cold, that would be important, but we are generally interested in epidemics. In particular, why do colds go epidemic in the winter? We know there must be some environmental change. Maybe it's a small change, since it only takes a small change in reproduction number to cause an epidemic. Then these controlled experiments might identify the main method of transmission. But maybe the change from summer to winter is a big change that swamps the effects we can measure in these bottom-up experiments.

I think Benquo's claim is that most institutions do want financial fraud. Most people don't, but a big constituency arranges to profit from it and to corrupt institutions. So his advice is aimed at the individual, not to blindly trust institutions.

Thanks!

I have a lot of quibbles about this, but going back to the main point, yes, this was a separate source of funding from stealing from customers and unlike Enron, not a rounding error compared to the first step.

via indirect channels like Alameda

What channels other than Alameda? If this was entirely about Alameda, how is FTT adding anything to the story above saying that they stole to give to Alameda? Who are they fooling other than their own internal accounts? The Coindesk article is very late in the story because no one saw the accounts before they tried to get a bailout from Binance, who wasn't fooled.

If a customer puts up FTT as collateral to short bitcoin, then FTX is confused about how much collateral it has. But this is the customer exploiting FTX, not vice versa! This is FTX exploiting all the other customers by falsely claiming that it has hedged risk. But this isn't what took it down. It did manage to largely liquidate shorts before they ran out of collateral.

Later you say that this is just one ingredient, but in the beginning you describe a rigid structure of compounding and shoehorn examples into bullet points. I think that structure is too rigid and I object to the examples. I think Theranos and academic Tessier-Levine fit it well. Indeed, for research, pretty much all people have is their reputation.

You describe Enron as a simple Ponzi scheme. I think this is just wrong. My understanding is that the main thing with Enron was a one-time shift from brokering energy today and owning the physical capital to deliver that energy, to brokering long-term deals that were purely virtual. By having unmoored forecasts, they could make arbitrary forecasts about the state of the company. It is not clear how much the senior executives knew they were doing this; perhaps they just accidentally designed poor incentives by allowing the salesmen to price their own deals and get an immediate bonus with no real-world feedback of how the deal turned out years later. You can certainly say that this is an example of extending credit from Enron's old business to its new business, but I think it was a one-time change and not repeated compounding of credit. It is true that the new business having no physical capital was able to scale very far on financial credit. (Enron is my hobby horse. "Everybody knows" that you're supposed to hold it up as an example, but it did 3 bad things and no one notices that other people are talking about different things. The main problem was that there whole business was mispriced. When the senior executives discovered this they something more like a Ponzi scheme, but that was just a couple billion, a rounding error. And the third was manipulating California energy prices.)

Similarly, FTX made a one time pivot from trader to bank. As I understand, it stole mainly to cover the trading losses and not to pay for advertising. If it had accepted the trading losses and wound down the trading, it wouldn't have looked very different from the outside. Trying to convert its credit for trading to being trusted as a steward might be an example of the credit conversion you're talking about, but not a compounded phenomenon. If that was central to its strategy, then maybe would make it hard to accept the loss and wind down that business.

What scale is "quickly ramping up their PR spending"? Theranos lasted 10 years. Enron lasted 15 years from the merger of much older companies. I don't think a heuristic about speed would identify either of them. It sounds like you're only talking about FTX.

The image is broken. I put it here. source

The denominator isn't fixed.

If you have a tool which does a good job at automatically generating a certain type of code, you should probably generate more of that type than you used to. This could easily balloon to 90% of your new code, but only replace a small amount of human coding. You seem to allude to this scenario when you talk about one-off scripts. Another example is automated tests.

During the last pandemic people were sometimes not allowed to wear reusable respirators, but required to drop down to N95. This is pretty strong evidence that people think they are weird.

Without disagreeing with the conclusion, I think this is a poor discussion of the pros and cons. The big con is weirdness points. That's basically why people don't do it. Probably if you articulate that it's easy to see that it's a win, but if you don't articulate it...

Yes, they're supposed to be more effective. I think this is emotionally difficult because it sounds like a Pascal's Wager. Yes, you should invest at the prepper margin, but how far? How many other points are comparably far down the list? What should I do before this? Here's a concrete point: if I have a beard and an N95, I will achieve more by shaving than by switching to a reusable mask. Shaving is pretty low cost (especially, conditionally shaving in the event of crisis), not a tradeoff with changing masks, but knowing to shave is important and competes with mental resources.

Here are two other advantages over paper masks: you say elsewhere that it is more cost effective than paper masks. If you are already committed to having a stock of masks, this makes it much less like a Pascal's Wager and more of free lunch. And I suspect that they are more comfortable, which is important for actually using them.