Flow-matching, the technique they describe is incredibly interesting. I studied it in the context of generative AI and found it fascinating. It’s so fitting that a technique that borrows from thermodynamics and uses Brownian motion would go full circle to solve for protein folding.
I’m not an expert, only dabbled in photogrammetry, but it seems to me that the crux of that problem is identifying common pixels across images in order to sort of triangulate a point in the 3D space. It doesn’t sound like something an LLM would be good at.
I just checked and the pm2 levels are as bad as they were when I visited pre Covid (50 μg/m^3). I was there for 3 months, in El Poblado neighborhood and the air quality was horrendous, I would never go back. The depression in which the city lays also keeps all the pollution in place for longer.
The author brings up a very good point about controlling the monetary policy which made me think about the impossible trinity. I have a very basic macro Econ 101 understanding of this, but it seems to me like the dials the central bank used to tweak things like inflation and unemployment are delayed and not precise at all. I think innovations in public ledgers along with things like automatic market makers can tighten the feedback loops so that hopefully one day we have such a high degree of integration that interest rates and inflation balancing can be tweaked in real time
Here are a few reasons why I called it despicable:
- Laundering money in poor countries directly hurt the citizens of those countries because the government doesn't collect the taxes it is entitled to
- Most of the money to be laundered comes from illicit activities: drugs, prostitution, human trafficking. I don't have any citations here for proportions, sorry.
I explained why I think this is argued in bad faith: trying to frame mining in a law framework for the single purpose of killing it vs something like: "ok, this is bad, let's try to fix it. here's how we can fix it"
I love Lawfare and listen to their podcast quite often, but the argument presented is so odd: 1) specifically calling for not introducing new
legislation to deal with this and 2) trying to fit mining into a bucket of legislation designed to kill it. This is not in good faith at all.
Money laundering is despicable, specially when it happens at high level. And we have to admit that Bitcoin, ZCash, Ethereum (Tornado Cash) and other cryptocurrencies are facilitating that. But we're getting better at tracking that. After all, transactions are all public and once you can label one wallet address, you can perform all sorts of graph analysis and learn a lot. Contrast that with trying to subpoena banks to follow money trails.
The other aspect of this is that, like it or not, banks are not playing by the rules all the time. See HSBC laundering almost 1 billion for Mexican and Colombian cartels. See Austria’s Raiffeisen Bank enabling former Ukraine's president to steal money with offshore accounts (https://www.occrp.org/en/investigations/former-ukrainian-pre...)
We should absolutely deal with this, but the first place to start is the Bank -> Cryptocurrency (and vice versa) transfers, or as they call them, the fiat gateways. You still can't pay for stuff with cryptocurrencies so ultimately, whoever steals or launders using cryptocurrencies, will need to eventually convert back into fiat money.
Yes, this is the kind of discourse I would expect to see around blockchain technology from a group of smart nerds. I can't argue with that, the oldest blockchain kind of failed to prove itself valid in significant ways.