I work for the Internet and am an advisor to @stripe. These are my personal opinions unless otherwise noted.
165 subscribers
Feb 17 • 22 tweets • 4 min read
Me: Ah a nice relaxing Monday where I can finally get some work done.
DOGE: Have you ever heard of checks?!
Me: %{*]% it.
Me: I don’t do partisan politics.
Twitter the Sumerian bird demon: Got it.
Me: Which is why I work in a painfully boring infrastructural field.
Twitter: Oh sure.
Me: That no one hates each other over.
Twitter: Yeah.
Me: So just writing the truth won’t summon a mob.
*curse starts*
Feb 14 • 17 tweets • 3 min read
Friendly neighborhood Dangerous Professional advice:
If you are ever at a meeting taking notes, and someone at the meeting expresses umbrage that notes are being taken of the meeting, and this is routine notetaking for this genre of meeting, …
… you should absolutely want to keep physical control of those notes, and you should prioritize that over social pressure you may perceive, and you should update very aggressively against the umbrage-taker as being likely up to no good.
Feb 10 • 9 tweets • 2 min read
Since the printing press, it was injurious to reputations to have something untrue said at scale.
We should have adopted more care about this after it became habit in society to Google someone's name and Google started weighting institutions highly.
And now we're standing on the precipice of another revolution in user behavior caused by a technological substrate almost unimaginable earlier: there will be a presumptively authoritative answerer of almost every question, in almost every pocket, and many places besides.
Feb 6 • 8 tweets • 2 min read
Per CoinDesk (and thanks @Bitfinexed):
Lutnik, to Senate: “Cantor Fitzgerald is not conducting continuous diligence on Tether’s financial statements, but I believe my [much stronger endorsement] statements were accurate when made."
Was wondering when walk-back would happen.
Lutnik, who was the CEO of Cantor prior to more recent adventures, previously attempted to dispel long-time doubts about Tether's... everything by vouching for them.
At the time, the open ended nature of the vouching struck me as out of the ordinary.
PDT stands for “pattern day trading” and the PDT rules were imposed after the Dot Com bubble to decrease risk to both small-dollar retail users and the brokerages that accept their business from day trading.
Jan 27 • 30 tweets • 6 min read
Well looks like I ended up with more relatable transpacific banking influencer content.
You either know if you’re in or you’re out for the rest of this thread.
It begins with the balance of one of my Japanese accounts suddenly going to zero, and me calling the bank.
Bank CS: *We do the KYC dance.* Oh really sorry McKenzie-sama but uh this is not an issue that this office can help you out over the phone. Monday morning please call your branch.
Me: Could you hum a few bars please.
Bank CS: I regret I cannot.
Me: To put my mind at ease prior
Jan 24 • 15 tweets • 3 min read
One of the reasons why PDF files that purport to be bank statements are believed to be reliable is that all serious adults understand that if you *$#(%( with one for advantage over a counterparty, you will be prosecuted, with very high probability.
A thing you'll often read in narratives of frauds (Lying for Money has several, and it is a rich genre in the bookstore of your choice) is that the fraudsters don't start out, on day one, doing this.
If you can have a sensible conversation with an engineer about their work, capitalism says you’re eligible for forty jobs that pay better than being a recruiter.
Why does the industry keep stepping on this rake? In part because if *has tried* to get engineers / EMs to sit down with a pack of resumes or LinkedIn profiles and sift through to find the highest priority candidates.
The first day you ask you’ll get takers. The second less.
Jan 21 • 5 tweets • 1 min read
“What choices would you make in a world where the great and the good comprehensively underrate not merely the future path of AI but also realized capabilities of, say, one to two years ago.” remains a good intuition pump and source of strategies you can use.
You wouldn’t think that people would default to believing something ridiculous which can be disproved by typing into a publicly accessible computer program for twenty seconds.
Many people do not have an epistemic strategy which includes twenty seconds of experimentation.
Jan 19 • 6 tweets • 1 min read
Correspondent: *situation* So how is it that my credit card at A shows a payment has posted but by checking account at B shows it as not withdrawn yet?
Me: That’s usual behavior for U.S. bank transfers. Each is an extension of credit, with various risk controls.
And now I’ll continue this with a character.
X: But the money is in two places at once. It cannot be in two places at once.
Me: Extension of credit creates money, yes.
X: Good thing this only happens for credit cards and not for checking accounts.
Me: Built into every account.
Jan 17 • 36 tweets • 7 min read
Seems like the SEC finally got around to fining DCG for lying through their teeth repeatedly in 2022.
I think this finally lets me cash in a very old hash. Most is very old news, but there's one discussion topic that is perhaps not public yet.sec.gov/newsroom/press…
So back in November 8th, 2022 I dropped this hash on Twitter.
That is not actually a book, but let’s say there is a very rich subgenre of advice, much of it written by and for very successful people, which says what that book would say, in some cases literally by way of extended D&D metaphors it assumes the audience will grok.
Jan 16 • 18 tweets • 3 min read
I sometimes get asked "Why don't you have a tip cup? I'd like to support you, but don't want to commit to a subscription." My macroed response is that I appreciate the gesture and suggest a donation to a local charity.
I do not take donations because I sell things to businesses.
Businesses do not customarily tip or donate to professionals. When they see someone with a tip cup, they think "Not a professional."
This is not something I want a business to think about me when quoting rates.
Jan 15 • 9 tweets • 2 min read
Matt Levine has a great entry today about the bank which motivated this tweet thread, which apparently ran afoul of the CFPB. He is… appropriately skeptical of their theory that this is improper in a way which should be illegal.
Two years ago I would have been pleasantly surprised that an LLM could math out what a QSR manager was and provide fictional examples. Today I’m pleasantly surprised one just extrapolated cultural commentary on why PMC values chefs more than QSR managers… without me asking.
Claude 3.5, FWIW. I am very new with Claude and have minimal system prompt or context it could be going on, unless it's cross-connecting with my public corpora.
The setup:
Jan 14 • 6 tweets • 2 min read
You’ll notice that in society we have many competing classes of prophets. The ones who actually have to be right about the future are despised, while the ones who are never scored on that continue being invited to the nicest parties.
Not at the nicest parties: insurance underwriters, prediction market users, conversion optimization specialists.
At the nicest parties: politicians, journalists, and people who publish in fields where replication is a thing you ask only of your enemies.
Jan 9 • 31 tweets • 5 min read
In this thread you have a crypto firm doing a soft freeze of a crypto VC’s account, an executive escalation, CEO involvement, and a swift rescinding of the freeze.
Pointing to it because a) all of this happens in the non-crypto side of finance b) unusually publicly documented.
And now for some commentary, which I will make broader than this particular firm/incident:
“Why don’t financial institutions flag certain accounts as belonging to responsible adults?”
In fact, they frequently do so, either implicitly or explicitly.
Jan 9 • 11 tweets • 2 min read
A conversation I have frequently about ACH speeds.
Me: The most important factor to understand is Reg E and the broader constellation that makes banks liable for consumer accounts being drained.
X: But other nations manage this without slowing down bank transfers.
Me: Which one.
X: Take India for example.
Me: UPI is an extremely impressive payments system and public/private partnership. What will a bank tell you if you are abused using it?
X: ‘LOL skill issue bro’
Me: Yes, really important to understand that American banks cannot do that.
Jan 9 • 39 tweets • 7 min read
“We” could change this but “we” have not prioritized changing it, for a variety of reasons. The “we” is more complicated than people often model “us” to be, and has crosscutting preferences implicated here, such that velocity increase is not good in all circumstances.
Many people erroneously believe that banks keep transfers slow primarily to earn interest income on “the float.” This is basically because they learned the financial industry from regurgitations of the career of Warren Buffett. In insurance, float is a very, very big deal.
Dec 30, 2024 • 7 tweets • 2 min read
This would have been straightforwardly true at a major U.S. research university in 2004, and I doubt facility with the command line has increased in the intervening 20 years.
Being able to use any source control system put you in top 1% (not exaggerating) and THAT I would predict has actually changed, thanks primarily to git and GitHub-centric OSS development and deployment.
Dec 28, 2024 • 9 tweets • 2 min read
I continue to think there’s a fairly titanic gap between the tech industry (as broadly understood) and someone in Columbus working as Network Administrator I at a mortgage company.
Often they end up talking past each other and/or assuming away existence of the other side.
One’s perpetually on the hunt for talent. The other’s perpetually an email away from losing their job, and the expectation they will take months to line up a new one.
