Thrift seems to have a really nice solution of providing only types that eventually break down into scalars. And only for exchanging data - it doesn't dare prescribe the rest of that. I liked working with it.
I had serious anxiety a few years ago about this, and ended up in Acceptance. I'm not happy ending up here, but I figure I got about 30 years left. That's pretty good, I'll take it.
Secret societies have an air of mystique around them, but often they're just social clubs with extra steps. People are social creatures, and having something to belong to is healthy for most!
Leonard Leo, who's had more influence over the selection of Supreme Court Justices than just about anyone, belongs to a Catholic secret society. It's not silliness or paranoia to worry about groups like these.
There have been real secret organizations, but they tend to be more like terrorist groups. The League of Free Jurists, the Irgun, etc. Ones that work as lobbies have to have some visibility. You can't stay completely behind the scenes forever and accomplish much.
The U.S. Chamber of Commerce is arguably one such. Its existence is not a secret, but its 1970s plan to make the US more friendly to corporations didn't become known until it had already worked.
I uninstalled most apps from my phone, and turned notifications off on the rest. If it's not a DM, I don't want to see it. My quality of life has improved.
An automated message arrives every time, one time. For this specific type of fraud where they are trying out many cards to figure out which works, getting an automated message is perfect.
Leaving the fraudster in the dark - excellent. Forcing the fraudster to call if they want more information - excellent. Both of these increase the time investment from the fraudster. They need to spend more time per card.
Cussing - doesn’t make a difference either way from an information perspective.
I worked with the fraud team implementing the security for a real time data ingestion pipeline at a major bank partner. I am a bit more informed on this than the average hn poster :)
It's literally less information versus directly letting them know. One message lets them know you know, and the other doesn't.
> Forcing the fraudster to call if they want more information - excellent.
But there's something else you're not taking into account, which is innocent people who trigger your fraud detection.
>Cussing - doesn’t make a difference either way from an information perspective.
Well it certainly lets the fraudster know you know. A legitimate customer receiving that kind of abuse would be pretty unusual, don't you think?
> I’m a bit more informed than the average HN poster
You’ve mistaken me for the average. I’ve worked in Integrity for a FAANG company and in FinCrime for a bank. I have a very good idea of how to mask information from bad people automating things. That’s literally all I’ve done for more than half a decade.
It doesn't change that this is basic logic. One yields less information than the other, and you didn't grasp that. Sorry if I came off as condescending. I don't really see a point in continuing this conversation.
>The FTC alleges that the company used claims that consumers were pre-approved and had ’90 per cent odds’ to entice them to apply for offers that, in many instances, they ultimately did not qualify for.
I believe CK has inside views to the models^ these companies use, and I wouldn't be surprised to find that 90% is actually very close to reality. However, I can also see why someone taking a hard credit pull would be very annoyed to be declined.
Also, Credit Karma gets paid for successful conversions, and maybe ad placement? It doesn't seem like misleading someone got them any profit.
This all around seems like a really weird thing to slap this company with - Credit Karma doesn't really directly profit^^ from getting this wrong, nor do their partners. Yes, Credit Karma screwed up, but to frame it as "misleading consumers" makes it sound a lot worse than it is.
I wonder if there is missing subtext or inside baseball that makes this all make a lot more sense. Regardless, that language does seem misleading, and I'm glad to see it be turned into something more accurate and informative.
^ they seem to have some sort of b2b platform ("lightbox"?) for letting their vendors import their models into credit karma. It's probably pretty powerful for a lender to change and simulate new model changes for targeting offers.
^^ pissing off your users while not making money is always a bad look
>Also, Credit Karma gets paid for successful conversions, and maybe ad placement? It doesn't seem like misleading someone got them any profit.
This is incorrect because misleading causes more people to apply. The people who are coaxed into applying through deception have a non-zero chance of turning into a successful conversion, which makes credit karma money. For instance, if people interested in a credit card, but they're not certain that they'll get approved, so they end up not applying. If there are 100 visitors in that situation, and credit karma lied to them, then they should expect to get 66% (based on the actual approval figures from the FTC) successful conversions (ie. profit).
I wonder if there is some selection bias going on. One hypothesis could be that the people who are financially more responsible (and thus are more likely to be approved for a new credit card) are also the people who are unlikely to apply for a new credit card just after seeing an ad in a website. Thus, credit karma ad clickers self-select to a lower success rate than what their model predicted.
This still seems a bit dubious—-90% of people who view the ad being qualified doesn’t mean in any way that 90% of applicants will be successful. If you have more applicants from that bottom 10% pool who click on the ad to proceed, then it would be statistically very easy to end up with a reject rate much higher than 10%.
It didn’t say 30% were rejected - it said on some offers as much as 30% were rejected. Again, the models can be totally accurate in aggregate but due to random variation or low sampling show as inaccurate on one specific offer, that is just how statistics work.
If a third of consumers _should_ have been denied because they were not credit-worthy, then there were no errors. You need to look at the false negative rate, not the absolute number of people denied for financial products.
There are no "false negatives." The models implemented by the card issuers are the literal source of truth as to whether a consumer does or does not qualify for the card. Anybody who applied and was rejected did not qualify by definition.
I think the misleading part was representing it as that person specifically had 90% chances, even though 90% may be aggregated probability and this person's chances based on personal history may be much lower. Given CK does have access to personal history, one might easily get an impression that the prediction is a precise calculation based on it and thus imply more accuracy than warranted.
I read it more as "using uncertainty" is a dark pattern and to stop it even if it's accurate. 10% of users a company with that many is a pretty significant impact to people.
I reread the article again, and it seems to me it was "pre-approved" that was the issue.
Regardless, it's good the behavior was noticed and stopped.
And yet it turns out the proportion of applicants not qualified was more more than 3 times higher than CK claimed, and their assertions that applicants were pre-approved turned out to be a flat out lie.
> wouldn't be surprised to find that 90% is actually very close to reality
I’d be blown away if this is the case. Credit models are carefully guarded. They’re also expensive to run. If Credit Karma could approximate the pricey model with open-source data, they’d have been bought by a bank, not a tax company.
> Credit Karma has a platform for running models against their dataset, and appears to buy everyone's daily data from Equifax
There are a lot of people doing this. For the aforementioned reason: a cheap approximation of a credit score is valuable. Before the financial crisis, VantageScore was developed by the credit bureaus to disintermediate Fair Isaac. This work continues, and Credit Karma is far from unique in its approach or data.
But CK isn’t predicting whether you will default, but whether another lender will regard you as a good customer (where default risk is just one factor).
That’s a different problem than simply evaluating credit risk, and is much less researched — so not the same as the highly guarded credit models.
> they posted a Ruby on Rails directory structure as proof of hacking them but the company does not have Ruby code
I think it's extremely suspicious, but often times breaches like this aren't through the core platform itself. For example, Equifax was a support site that was hosted and built separately from their main platform.
This whole thing does smell like BS to me, though as well.
There are massive use cases for it, but not at the people level. Low latency tasks such as edge AI classification, IOT interaction, and game streaming are all currently limited to WiFi only.
> Once that infrastructure is built people will use it.
I would argue that that statement is true because it's talking about 5G whose features vastly outcompete LTE in many areas. I suppose there is a chance people won't use 5G but I think it's really unlikely considering the standards already been adopted and being used by many big name players.
All of this was to say that 5G has applications even if they might not appear to the OP and that it's only going to be used more once people can actually access 5G technology. It's still in the early stages even in areas which claim that are on 5G for the most part its 5G NSA mode where the backing core network is all LTE still. I also feel as though OP was really talking down 5G trying to bring nonsensical technically problems and unproven medical problems that have no evidence.
You yourself pointed out several applications but your quoting the very advantage I was talking about which was latency. Which we both agree is extremely beneficial but the over arching point is that we don't know all the things that will benefit from 5G because we have not observed them and while IoT, AI, and streaming will absolutely benefit the benefit does not end there. There absolutely will be more areas that benefit which is what I'm trying to communicate
It's not in the general, constant sense that you have to worry about, but rather in specific applications. When you're out in public, "they" will have 3-meter-accuracy, more than enough. But when you're in stores, and malls, and other venues where UWB is set up, then that's where real problems begin. They'll be able to track which advertisements you linger around, and which sections you visit.
It's going to be a whole new category of passive location tracking.
There are companies doing this already, but with their current CCTV systems. You don't even have to buy equipment, just the software, and you get e-commerce metrics for brick and mortar stores
My opinion is that in-building tracking doesn't have to be an issue, and the people who care don't linger watching advertisements in Malls. More power to someone who finds a way to use that data to make buildings like Grocery stores more efficient, like getting room temp products first and frozen things last during the walk.
