>Moving money is one of the oldest human endeavors and unlike a lot of topics in software development, it's an old profession with a huge body of accumulated domain knowledge. We shouldn't be so quick to dismiss it.
I would say that, like many fields, the existing domain knowledge in the relevant professions (accounting, finance) is incredible. But CS and the sort of expertise common in the Bay Area does have a lot to contribute to the tools used in those professions. I'm a lawyer and a developer working at a large public account firm, and my experience has been that the profession has a lot of experience recognizing the myriad ways to identify, measure, and disguise value via financial transactions, but very little expertise in using computers to do so.
Yeah, you're not gonna fool us by withdrawing $100k from your account and hiding it under your mattress or in a crypto wallet. Everyone in the field knows to look at the economic substance of a transaction over its nominal form. But most of us can't figure out how to open a CSV file that's too big for Excel, so there's plenty of room to help us with managing lots of data and extracting useful information from it.
>Case 3: money laundering and other types of financial crimes. There are entire companies who build software to detect money laundering. It's a well-understood problem.
I would disagree with you here. Money laundering is an extremely challenging problem, not least of all because money launderers are extraordinarily inventive and constantly shifting tactics. It's a lot like infosec. No matter how advanced your defenses are, a ton of hackers are always looking for a way through and any sufficiently complex system is impossible to secure 100%.
Not the OP you responded to but there's an important distinction between "well-understood problem" and "solved problem" and I believe this specific choice of words was deliberate.
Ok, that's a fair hair to split, but I would also disagree with your more explicit characterization. I think the recent disclosures via the Panama and Paradise Papers demonstrates that the extent and nature of money laundering remains poorly understood by regulators, academics, and the public at large, though perhaps there are insiders (meaning actual money launderers) with deep insight into the activity. We may know a lot about some basic red flags, common motivations for money laundering, common funding sources, which banking markets are grey or black, etc., but that doesn't mean we understand anything about how it is actually being conducted on a day to day basis. If we did, it wouldn't be so pervasive.
I would say that, like many fields, the existing domain knowledge in the relevant professions (accounting, finance) is incredible. But CS and the sort of expertise common in the Bay Area does have a lot to contribute to the tools used in those professions. I'm a lawyer and a developer working at a large public account firm, and my experience has been that the profession has a lot of experience recognizing the myriad ways to identify, measure, and disguise value via financial transactions, but very little expertise in using computers to do so.
Yeah, you're not gonna fool us by withdrawing $100k from your account and hiding it under your mattress or in a crypto wallet. Everyone in the field knows to look at the economic substance of a transaction over its nominal form. But most of us can't figure out how to open a CSV file that's too big for Excel, so there's plenty of room to help us with managing lots of data and extracting useful information from it.
>Case 3: money laundering and other types of financial crimes. There are entire companies who build software to detect money laundering. It's a well-understood problem.
I would disagree with you here. Money laundering is an extremely challenging problem, not least of all because money launderers are extraordinarily inventive and constantly shifting tactics. It's a lot like infosec. No matter how advanced your defenses are, a ton of hackers are always looking for a way through and any sufficiently complex system is impossible to secure 100%.