- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
You must log in or register to comment.
it is the right moment for everyone who builds systems of control to reflect. It’s the right moment to pierce those layers of abstraction that allow you to get through each day, and question why it’s so financially lucrative for the system you’re building to exist.
Because there is no abstraction as leaky as a man waiting outside your hotel at 6:45 in the morning with a gun and murderous intent
Amazingly written article, last line giving me chills
This one also a bit chilling:
A data scientist can tell most of someone’s life story given their zip code, and we try not to think too hard about why that’s always the most predictive feature in the model.
Elvis Presley is able to answer that one, and he’s been dead for almost 50 years. Makes you understand why his fame is so lasting, unlike the teen heartthrobs of later decades.
Some credit should go to the person who actually wrote that song, Mac Davis.
deleted by creator
Someone please make a bot that posts articles from this author.
Most of his stuff is about product/project management, so this was quite a break from the norm.
Dude can write!
Can’t you subscribe to substack writers/sign up for email notifications?
I could, I guess. I miss when RSS was the norm.
It’s the right moment to pierce those layers of abstraction that allow you to get through each day, and question why it’s so financially lucrative for the system you’re building to exist.
I’m glad someone said it because this thought popped in my head yesterday. Been thinking about the consequences of my system, and really if it brings benefit to the users, but also who it affects indirectly.
So far, I’m ok with it. There is part of it that adds some safety for the business, the users, and people affected indirectly. But it still has a profit motive and that’s the uncomfortable part.
Edit: I should clarify that I’m talking about my software system. Not the healthcare system in the U.S. like the author is. It’s nowhere near as lucrative as making money off of people literally suffering from life. But the author mentioned how the CEOs see numbers not people. If the numbers my system collects ends up hurting people, that’s what I was reflecting on.
There is nothing wrong with making a profit. People have to be paid, after all, and that includes the ownership who put the money at risk in the operation to begin with. The problem is when making a profit becomes the only motive.
Every company is established with the purpose of offering a product or performing a service that makes their customers’ better or simpler. If is successful, it grows from nothing to something in a relatively short period of time. Then it gets the attention of the Investor Class, who shovels money into it with the expectation that it will sustain that growth. Now, the focus is on Building Shareholder Value, and the customer is seen as a necessary evil toward that goal.
The worst thing that ever happened was when we decided that public corporations had a duty to maximize shareholder value above everything else. It renders all those mission and vision statements irrelevant. No matter how much the CEO says the firm’s goal is to make the world a better place through selling stuff, we all know it’s a lie. Their goal is to enrich tthemselves, at our expense.
If the people putting money in deserve to be paid for that money, it can be treated as a fixed term loan, with an established interest rate. That makes it a business expense.
Profit is what’s left over after everyone is paid for their work, and the costs of materials, housing, and maintenance - invluding the maintenance of debts - are covered. It’s either what you’ve over-charged your customers, or underpaid your employees.
And that’s wrong.
The most amazing part is not even that long ago, everyone agreed this is how it worked, even the business owners. I remember recently watching the 1923 silent film “Safety Last!” starring Harold Lloyd. I was very struck by a particular scene in the film where the owner of the store Harold Lloyd works at says:
“I’d give $1000 for a new idea that would attract an enormous crowd to our store. Something is wrong with our exploitation! We simply are not getting the publicity that our position in the commercial world calls for.”
This character is not presented as some kind of villain or saying something wrong. He’s just talking about how everyone understands business to work, by exploitation, which has always simply meant taking advantage of some kind of opportunity, even when people like Marx talked about it.
I added an edit to clarify my reflection.
It’s one of the reasons I couldn’t go into the defense industry. Not just working on weapons that are deadly to enemy combatants and innocents; but making profit off of doing so.
If there becomes a point in my career where it’s clear that my work doesn’t make things better, then I know I’ve made a mistake.
I know a couple of people who had jobs working with a defense contractor. One of them justified it by saying he doesn’t actually work on missile projects, his department does as air traffic control systems (and sure, there’s nothing objectionable about that). But they both knew what their company did and it certainly did make it an uncomfortable place to work. You can tell yourself that these weapons only get used on bad guys, but I think the more you have to tell that story, the harder it is to hear yourself.
They’ve both moved on at this point.
Everything I got offers for were weapons platforms, fire control systems, or guidance.
The exact same time I was going through these a cruise missile in Yemen hit a school bus full of kids - obviously one American made sold to the Saudis.
My kid wasn’t in school yet, but I looked at him, shook my head and said nope. Can’t do it. Won’t do it. Turns out I do value things other than fun projects and money, and by a huge margin.
If the USA keeps putting off single payer. The first thing that should happen is all health insurance companies be required to become non-profits and cap CEOs and other c levels pay at a certain % of the lowest paid workers.
profit is the excess, the infinite growth
And that’s one thing I like about the projects I work on. Nothing I’ve built has been directly responsible for profit, it has just supported other profit centers.
My current project helps us sell our main physical product by making the supporting software easier to use vs competitors. Yeah, the features highlight the benefits of our product vs competition, but the user is free to use any competitor they want, and we even have an open-ish API so they can make their own interface. We charge for it, but it’s far from turning a profit since the main point is to be something our sales team can bundle with the main product.
We build software for reports, simulation, design, etc, and the entire goal is to be useful, not extract profit. We charge for computationally heavy features, but that’s more to prevent abuse (i.e. keep costs reasonable) than anything.
My company also has direct competition and who has decided to go with the lockin approach, and customers seem to appreciate us as an alternative. The business itself isn’t particularly ethical, but it’s necessary, so it helps me sleep at night.
That said, our end goal is to replace good (but dangerous) jobs with automation. and that will be complete once we plug the leaks in our abstractions, and that’s a little sad. So it goes I guess.
First time I’ve heard of the “Leaky Abstraction” concept, makes a lot of sense. Good metaphor too.
The Hacker News crowd uses this phrase every other sentence so it was almost humorous to see it used here. I thought this was a shitpost
Yup. Fortunately, I’ve been away from HN long enough that I didn’t immediately avoid the article, and I’m glad I didn’t.
I first heard of it from Joel Spolsky’s blog and wikipedia also credits that article with popularizing the concept. In it’s original formulation, it was based on remote procedure calls being hidden in APIs. Because a remote computer call has all these limits of latency, packet/info loss, and possible connection loss, it is impossible to make a perfect abstraction that allows the programmer to treat the remote call as though it were local. The reality the abstraction tries to hide “leaks” in those fundamental limits.
All of contemporary global society is such an abstraction; that’s one of the principles of post-modernism. When you buy clothes online an entire invisible work force of shippers, manufacturers, resource procurerers, and more lies beind each article of fabric.
Pressure from climate change, tariffs, global war, and more are straining the foundations of society and the comfortable abstraction is starting to crack.
We need these abstractions though, we love them, and we should! The problem isn’t abstraction itself, we couldn’t survive without that. The problem is in how we build these abstractions, the problem is that we create unsustainable and irresponsible abstractions and then present them as legitimate options.
But we can’t throw out abstraction entirely. It’s like, if a person dies in a car crash, we should ask some questions, but you don’t ask “should we really have cars?”, instead you ask “should that person be driving a car? And is that car safe to drive?”
I didn’t recognise the ‘10 seconds of human consideration’ claim. I found this ProPublica report on it: How Cigna Saves Millions by Having Its Doctors Reject Claims Without Reading Them
There a former Cigna doctor says “It takes all of 10 seconds to do 50 at a time.” They claim to have seen documents for a two month period that put the average at 1.2 seconds per review. That’s using a specific review system that processed 300,000 claims over that period.
They don’t mention if there were other claims processed with different methods but still, the OP article seemed to be generous with that claim.
I don’t know what the “90%-error-rate AI” claim is about though. It’d be nice if the sources were actually cited.
Because there is no abstraction as leaky as a man waiting outside your hotel at 6:45 in the morning with a gun and murderous intent.
A leak like that is the type that keeps leaking. The system needs to change, and it eventually will, doubtfully but hopefully peacefully.
Good article
Damn good article
More precisely, medical business strategy is a leaky abstraction; the assassination is the leak.
One of my two major projects is a long-term reporting system on a sustainability initiative to help managers figure out whether their unit is compliant (definitely not for control, of course, nooo… though they are expected to talk to their respective subordinates if their results deviate too much, which probably filters up the chain when a given higher level breaks down their subordinate units’ figures).
Probably a PR push (I swear, if I ever see a figure calculated by my model in the newspaper, my impostor syndrome is gonna thoroughly shit my pants for me), maybe a move to get ahead of competitors in the face of legal stuff I’m not in the loop about, but doing the right thing for selfish reasons is still the right thing.
The other project… Well, I’m trying to push for measures that prevent user-level evaluations, but it’s a kind of corporate limbo right now. I’m doing my best, but that’s not a whole lot in this case.
