@The_Nybbler's banner p

The_Nybbler

Does not have a yacht

8 followers   follows 0 users  
joined 2022 September 04 21:42:16 UTC

				

User ID: 174

The_Nybbler

Does not have a yacht

8 followers   follows 0 users   joined 2022 September 04 21:42:16 UTC

					

No bio...


					

User ID: 174

The Republicans would rather lose forever than tear apart the country that way.

The use of the classified cover sheets in that photo does many things

  1. It provides a lot more visual impact than just classified documents with markings.

  2. It gives the impression that it would be obvious to anyone who casually looked in the box that it had classified documents. This is important because "knowingly" is an element of some of the charges.

  3. It effectively substitutes the FBI's CLAIM that the documents were classified for the actual evidence of classification.

  4. Since the classification markings on the pre-printed cover sheets didn't have to match those on the documents, it provided the impression that the documents had perhaps a higher classification level than they did. For instance, the NPR story claimed one of the cover sheets said "UP TO HCS-P/SI/TK", leading them to believe Trump had documents related to HUMINT. I thought at the time this was odd, you don't put "UP TO" on your caveats. But it makes perfect sense for a placeholder that might be used for a wide range of documents you might find. And given that, there might well have been no HUMINT at all; the placeholder is not evidence.

  5. Since the narrative accompanying the photo in court filings did not reveal that the cover sheets were added by the FBI, it constitutes an attempt to prejudice and/or mislead the court (as well as the public)

Last I heard, the fights over schooling worked out okay for Republicans.

Nope. As soon as the trans stuff disappeared from the headlines, the voters promptly forgot and voted the same school board right back in.

Departments will tend towards policies that let them do it, like stacking all the product in one spot. But does that make the drug bust illegitimate?

If the gold-plated guns were actually props (not recovered in the bust), it at least risks poisoning the jury pool. And that photo wasn't actually just a publicity photo -- it was included in a court filing by the Justice Department, so it also IMO constitutes an attempt to prejudice the court.

The part you're forgetting is that if Ford has to insure against all those accidents then the driver doesn't. The up front cost to the consumer may be more, but it's effectively prepaying an insurance policy that lasts the life of the vehicle.

Yes, but I claim the per-accident cost for Ford will be more because liability is not limited to policy limits + net worth of driver, and because plaintiff's attorneys and juries will know this. (Not to mention adverse selection of bad drivers into Fords, but that doesn't apply to the self-driving case)

The "placeholders" are part of their strategy of trying the case in the media, e.g.. Not just the visual impact of the cover sheets, but media people (including NPR in that article) using the caveats on the placeholders (provided by the FBI) to show what a horrible thing Trump did.

If you regulate mass-produced end-user consumer goods, you will destroy the culture of innovation in that sector, yes. But that's what you want, you've said so yourself; you explicitly want to change the culture of the outgroup you have that consists of software people who refuse to color within the lines.

If Ford was fully liable for any accident in which a driver of a Ford vehicle was found at fault, but this did not apply to any other vehicles, how much more do you think Ford vehicles would cost than all those other vehicles to cover that liability? I expect it would be at least an order of magnitude; being involved in an accident with a Ford vehicle would be a potential lottery-winner (regardless of who was at fault, and that's often muddy). And I think that's true even if from some nonexistent objective observer's POV, the Ford driver was never actually at fault.

In case there's any question left about the press's lack of objectivity, the CNN article you cited -- article, not editorial, not column -- contains this bit:

The move by Cannon is a significant win for the presumptive 2024 Republican presidential nominee. The proceeding will give Trump and his attorneys a platform to air unfounded theories about the prosecution, including the accusation that it is politically motivated.

Ok, I think we've made progress. It is literally impossible for the culture to bother taking the most basic steps to make the billions of devices on our networks not trivially hackable without causing some folks like you to shut down and stop being open to new ideas.

No, it's literally impossible to make a regulatory culture without shutting down new ideas. Perhaps there's some other way to get the result, but you can't have a culture with both properties. Once you put the commissars in place, initiative declines sharply, and that's unavoidable.

I guess we'll find out.

We've already found out in other areas. We just refuse to learn the lesson.

No, you're asking for an impossibility. You can't have one culture which is both open to new ideas and dedicated to checking boxes.

Once you've changed your culture from "building cool stuff" to "checking regulatory boxes and making sure all the regulation-following is documented", you've already done a vast amount of damage. Even if the regulations themselves aren't too onerous.

Yes this is a trivial problem to solve. We already have a massive auto insurance industry. Everything looks like self-driving cars will be safer than human drivers.

The auto insurance industry works because there's a relatively low cap on liability per accident (after which they stop paying out and the driver is on the hook, but the driver is an individual who likely doesn't have much). Once the manufacturer is on the hook, that cap is irrelevant.

Is this the bitter voice of experience of someone who has worked on software for the financial industry?

Not financial, but the meetings and the acronyms (though not the specific paragraph numbers) are real.

In my experience, companies that operate in compliance-heavy industries that also have hard technical challenges frequently are able to retain talented developers who hate that kind of thing, either by outsourcing to Compliance-As-A-Service companies (Stripe, Avalara, Workday, DocuSign, etc) or by paying somewhat larger amounts of money to developers who are willing to do boring line-of-business stuff (hi).

This works when the regulations target parts of the product that can be isolated from the technical challenges, but not (as in e.g. aircraft) when they can't. But I can understand the bitter envy towards software people of someone who is in a field where a good year means finding that you can tweak the radius of the trailing edge of the winglet by 1mm and save an average of a pound of fuel in an Atlantic crossing and only have to go through an abbreviated aerodynamic design review.

Insurance helps individual drivers because they can pool their risk with all the other drivers. A self-driving car company selling a sufficient number cars may as well self-insure. And yes, the expected cost of liability would be baked into the cars in either case, but I expect if they got it right, self-driving cars would be prohibitively expensive. If they got it wrong they'd go bankrupt when they big verdict came up.

Do you think the cost of self-driving car insurance would be higher than human-driven car insurance?

Human-driven liability insurance doesn't cover everything. Losses are limited. If you start a chain-reaction collision and kill a whole bunch of people, your insurance company will pay our to your maximum, and you'll lose everything you have and have to declare bankruptcy. Maybe you'll go to jail or commit suicide, but either way your victims ain't getting anything more. If a self-driving car does the same, the losses are limited to the value of the self-driving car company, which is likely far greater than any individual. And the company has far more exposure. And the plaintiffs and juries know the self-driving car company has much deeper pockets than an insured driver, so I expect you'd see more lawsuits per incident and higher judgements.

"well you want to turn software into an over-regulated morass similar to what aerospace / pharma / construction have become".

In support of this interpretation:

https://www.themotte.org/post/995/culture-war-roundup-for-the-week/210060?context=8#context (whole thing)

https://www.themotte.org/post/995/culture-war-roundup-for-the-week/209894?context=8#context ("Maybe their little subculture will change.")

https://www.themotte.org/post/995/culture-war-roundup-for-the-week/209881?context=8#context ("coloring inside the lines")

The inevitable increase of regulation once a regulatory framework is in place is part of it. Another part is that merely having a regulatory framework transforms your industry from "building cool stuff" to "checking regulatory boxes and making sure all the regulation-following is documented". Once the principals know they're going to be put out of business or go to jail for not following the regs or having the docs for following the regs, the whole development process is going to get bureaucratized to produce those docs. This both directly makes development much slower and more tedious, and drives the sort of people who do innovative work out of the field (because they didn't get into the field to sit in meetings where you discuss whether the regulatory requirement referenced in SSDD paragraph 2.0.2.50 is properly related to the SDD paragraph 3.1.2, the ICD paragraph 4.1.2.5, and the STP paragraph 6.6.6, which lines of code implement SDD paragraph 3.1.2, and to make sure the SIP properly specifies the update procedures)

I don't know. All I know is the consequences. Once the cars become popular enough, a self-driving car company is basically going to be mostly a legal company, defending (or settling) lawsuits in all 50 states involving its cars. And that's even if its cars are perfect and never cause accidents, especially since the car company is going to look like "deep pockets" to plaintiff's attorneys and juries. The cost of all this legal defense is going to increase the cost of the cars by a ridiculous amount, and the more cars there are the more of a chance of a "reverse lottery" where a self-driving car is involved in an accident that kills a busful of kindergartners and is found liable for more than Alex Jones even was. As long as there's a fairly small number of cars they can play the odds, but a liability regime which involves a car manufacturer in every major accident one of their cars is involved in will kill the whole thing.

Liability also doesn't come into play until the suit is underway.

The expenses start immediately.

And yes, I predict that if actual self-driving cars become more common, either we will see limits on liability or the companies will be driven out of the market or out of business.

Unfortunately, if manufacturers of self-driving cars can be sued for all accidents in which self-driving cars are involved (the "caused" part doesn't come into play until the lawsuit is underway), self-driving cars are essentially banned. The cost of covering that liability is staggering.

So while people might have supported the ADA if it was 1% of the budget, they might start getting pissed at the program when it balloons up to 10% of the budget and a bunch of reverse lottery sob stories start showing up in the news. And suddenly instead of 10% or even 1% of the budget, you get 0% for your cause and no one trusts you with a 1% allotment cuz they will all remember the horror days of 10%.

Except that's not what happens. Your program lasts forever because it sounds good to the normies and has strong built-in constituencies. So there's no incentive NOT to do this; if you do it you win.

You're trying to use "rational discussion" as a cudgel to get people to accept your conclusion that a regulatory framework is a good idea and the rational thing to do is argue over the details. And you're doing it clumsily.

Gay marriage was on a slippery slope down to all the trans stuff we have today. I don't know if the slope ends before dog marriage. Not sure what that has to do with a regulatory framework being a slippery slope towards the death of innovation.

The draft not working very well in Vietnam was why we got rid of it, though.

We got rid of it because we withdrew from Vietnam