discoursism

> The increased availability of credit, and the increased willingness of markets to give credit to risky borrowers has forced people to extend themselves to the max in order to compete for social status.

This is optional, though. If you opt in to it, that's somewhat on you. Granted that our social systems could be improved so that natural human tendencies push people toward right actions. However, I think the injustice is far more severe in cases where there's no out that's within a person's power to reasonably grasp.

And "the girls at the next table in the bar," don't know about your friend's car or house, unless your friend is an insufferable jerk who carries pictures of his ride in his wallet. And they likely don't care that much either way.

A perk is not a perk because of the ease of providing it or its cost. It's a perk because it's a benefit not part of core compensation. And it is true, none of the other tech giants have downtown HQs (although plenty of Google satellite offices are in urban centers, at least).

Of course. What do you expect?

Take "as individually as possible," to mean, "as individually as practically possible," rather than, "as individually as theoretically possible," and you'll have a better understanding of what I was trying to communicate.

No, I still value risk smoothing. I know a lot about my driving skill, but I don't have the ability to make perfect predictions about all possible future losses.

> I mean, that would asymptotically approach a prediction of your personal future losses, and thus your premiums would asymptotically approach your future losses, and thus what's the point of giving your money to an insurance company only to get it back later?

If we could estimate that well, risk would no longer be a real thing, and then there would be no need for insurance at all. However, you can't just take ideas to the absolute theoretical limit and expect the results to come out making sense in the real world.

I can't say for certain, because I wasn't involved in the decision. ;) My gut reaction is that it only ends up being seen as OK because the group being penalized has historically been at an advantage in most other areas financially. I am aware that some European countries are moving to do away with charging by gender.

I'm not a lawyer, but I'd say if you make a good faith effort to prevent that eventuality, it would probably at least mitigate liability. That's a great question, though.

Let me ask some clarifying questions, because I think I understand the broad strokes of what you're saying, but I'm not clear on some details.

OK, so I find that among Black users, Hotmail is associated with a 1.05 relative risk to the baseline, and among white users, 1.15 relative risk. How do I then apply this knowledge to an incoming user of an unknown race with a Hotmail account? Do I give them the population weighted average of of the relative risks among all my buckets? What if I do know their race? Do Black users get assigned 1.05 as much risk and white users get 1.15?

Also, what if some of my buckets aren't full enough to get a high confidence of the relative Hotmail risk in that bucket? Let's say I don't have enough gay Hispanic women in my cohort. Do I just drop them from the analysis and hope they're similar to the other populations?

Personally, I'd rather car insurance were tailored as individually as possible. If I could get that option without my data ever leaving Google's silo, I would do it in a heartbeat. Then again, I know I'm very low risk compared to most people.

In general, I do think it's good that people pay insurance based on the real risk of loss they pose. I'm fine with this in the case of umbrella insurance, car insurance, home insurance, life insurance, etc. The only exception I have is health insurance, because I don't think people should die of curable illnesses, just because they can't afford treatment.

But if someone poses a high risk of causing a car accident, I prefer that they not be able to drive. If someone is likely to burn their house down (maybe they're into basement welding), they probably should have to pay more to insure it. Insurance companies know that many people would prefer to subject themselves to the panopticon to save money, which is why some of them are starting to offer the option of installing a speed tracker in your car, among other things.

I recognize that this is not everyone's preference of course, but you asked, so I thought I might give you some perspective from someone who thinks differently from you.

I'd say there's a one really important difference between being Black and using Hotmail, and that difference is the same reason why one is a protected class and one isn't.

The real concern here isn't discriminating against people with Hotmail accounts. On the surface level, there's nothing wrong with that. This does become an issue if the property of having a Hotmail account is used as an avenue to discriminate against people based on their membership in a protected class -- a sort of actuarial parallel construction, if you will. At that point, we do have a bit of a problem.

> We need regulation in place to stop any punitive decision making, public and private, which can be found in court to be based on correlation instead of causation.

So, young men could not be charged more for insurance than young women? In fact, young people in general would have to get the same price as everyone else? People in New York would have to pay the same as someone in rural Alaska? Someone who's been driving a year gets charged the same as someone who has been driving fifteen, etc.? Even being in an accident doesn't cause you to get in another accident. It's just correlated with that outcome. Throw out the entire actuarial aspect of insurance? These "punitive" charges are all based on correlation. I don't see this happening.

----

That said, as a non-statistician, I'm interested in what kinds of tests can be run on signals to verify that they're not proxies for selecting and discriminating against protected classes. Is there a way to modify a signal using its correlation with membership in these classes such that you can marginalize the attributes you're not allowed to discriminate based on?

For example, it seems reasonable to believe that hotmail use is correlated with age, race, gender, etc. As an insurance company, I would have information on hand to inform me as to the extent of this correlation, at least among my customer base, and I would in turn know the correlations of those factors with risk. Could I somehow remove the racial, gender, and age components of the Hotmail risk signal to obtain a signal that only conveys the portion of the risk correlation that is not based on those classes? If so, what is that statistical technique called?

That's very true. And if they are not confident about their predictions then it might be difficult to push back when people fraudulently claim to not have purchased something.

Even with actual emotional support animals, I don't understand why anyone needs a seventy pound dog as an emotional support animal. Is a twenty pound dog insufficiently supportive? Are there a certain number of support units per pound?

In everything there's a balance between personal freedom to do whatever you want, and the accommodation of other humans who may have conflicting needs or preferences. When there are such conflicting needs and preferences, I feel it's incumbent on the party imposing unusual requirements on others to minimize their imposition. When it comes to selecting an animal to accompany you on flights, this entails choosing an animal that is likely to fit comfortably, not be unusually odorous or violent, and otherwise be relatively inoffensive to one's fellow passengers, at least to the extent one is able. (I understand that only certain breeds of dog are suitable as seeing eye dogs, and for that I will gladly make an exception, as these animals are usually exceptionally well-trained and are also almost necessary for their owners to function).

That and a reality inversion where this is in Amazon's or Google's best interest despite the massive blowback it would cause.

The site you linked to is satire. It is not a thing that actually happened.

I'm not referring to facial recognition. I'm saying if you shoplift, they are going to be able to provide the cops with video evidence, and they will know who you are.

But Alexa and Google Home aren't always on listening devices in your home -- at least not any sense that matters. They are always-on, and they do have microphones, but they are not recording every conversation and sending it to the mothership. You need to affirmatively request them to listen (via OK Google/Alexa/whatever hotword) before they are listening in a sense anyone should be concerned about.

Do you think they would have opened it if they didn't have a pretty high degree of confidence that it was working? It's been in beta for quite some time.

There are better places to steal from than those with massive numbers of cameras pointed at you from all directions.

> Can’t agree more. OP argument only holds if “value” is an absolute and objective concept and if everyone in the company is measured by it.

OP's argument is not a "true or false" proposition. It's a model, and as such it has some degree of correspondence with reality. That correspondence may be low or it may be high.

I think it is a useful model. It has a high degree of correspondence to what I see at one of the big internet tech companies where I work. People similarly situated would be well advised to consider this model when deciding how to spend their time, at least if their goals involve getting paid more.

If this doesn't match your experience, consider that this may simply be that your firm does not work the same way as the author's. It doesn't make you or the author wrong; you're just looking at different things.

And I also think it's fair for you to use the shorthand, "The constitution says ..." as a stand-in for this. But there seem to be some people in the thread who are not too familiar with the American legal system, and I don't want those folks to be confused into thinking the document actually says anything directly about corporate person-hood and free speech. Not trying to put you on the defense or anything.

I set up my 401k. That was work. I've also done the mega-backdoor Roth IRA last year, and will do it again this year. Am I a bad man?

The constitution as interpreted by five out of nine sitting judges at the time the question was last ruled on. It's reasonable to think the constitution says otherwise, although you are correct as to the law of the land at the moment.

This article is full of good observations. A shorter way to say it is that the distance between Steam and "perfect" is not very large, and there isn't enough margin to be had to justify the cost of building something to close that gap. And that's assuming you can identify the features that would bring you to perfect from where Steam is.

I cannot even imagine a game distribution product that would get me to switch from Steam, unless there were Very Important games on it I couldn't get on Steam, or unless the prices were at least 10-15% lower.

Punishment (retribution) is only one of a number of reasons to incarcerate someone (the three others normally cited are deterrence, rehabilitation, and incapacitation). I wish we had better metrics for the probability of recidivism. If we knew with a high degree of certainty that a given murder was a one time thing, we could let the perpetrator out very early. On the other hand, if we know that the perpetrator is likely to reoffend, or if we could show that light punishments had less of a deterrent effect, or if society just wouldn't stand for a lesser punishment, we could keep them locked up for longer. Unfortunately we just don't have that level of introspection into the human psyche.

In the sense that those processors can compute things which are computable, yes, they are a real alternative. The correct question is, "How much worse would the situation have to be with Intel before the cost of these alternatives was less than the benefit of switching to one of them." The answer, I guess, is: "much worse than things are now." However, if you're running at billions in revenue per quarter, you can afford to spend a few million to limit your downside risk by keeping yourself apprised of what the costs would be of moving to another platform.

Do you know that, or is it your conjecture? There are a lot of other possible reasons to use $500 as a limit. For example, preventing people from structuring withdrawals at the ATM, preventing an accident from producing too much cash, limiting the frequency with which ATMs need to be serviced, and limiting the downside of unauthorized withdrawals.

It is true in some sense that the banking system is eventually consistent, because commands can be issued asynchronously. For example, you can issue a check, and expect it to be drawn in the future. However, that does not necessarily apply to the ledger itself, and given what I understand about banking technology, I'm somewhat surprised if it is truly as you describe.

If it is indeed a conjecture of yours, I urge you to make it more clear in your comments when you are speculating and when you are speaking with authority.

Not to my understanding. They were already applying FDO to all of their performance sensitive programs. They found that after FDO, the retpolines and the other work done to mitigate meltdown and spectre had negligible impact. This is because FDO replaces indirect branches with direct branch hints, so only in uncommon cases do you even have to execute the retpoline. That's my understanding, at least.

Indeed. Google's published data suggests negligible losses if you are using FDO, which you must be if you care about performance.

If you weren't considering Intel alternatives before this, I'd argue that's a real failure of imagination and risk management. I'm sure some of the really small cloud providers weren't, but all the big players keep tabs on the path to and pain of migration, at a minimum. Just because they weren't actually using PowerPC/ARM/AMD does not mean they did not know how.

Paul is a boss. This is just one of his many amazing accomplishments. He's kinda the guy you go to if things are slow and you need some help to make them really ridiculously fast.