Originally published at: http://boingboing.net/2017/02/22/too-big-to-fail.html
Originally published at: http://boingboing.net/2017/02/22/too-big-to-fail.html
How do we design regulations that are harder to cheat? Easy, follow 45’s example. He has already signed EOs to roll back regulations for coal and banking, not to mention making every effort to gut regulatory entities such as the FCC, EPA, USDA, and others. There’s also his ridiculous “two regulations must be eliminated for every new one” EO.
So, the solution is simple: with no regulations, there’s nothing to cheat.
Sitting on the precipice of a huge advance in AI should make your skepticism even greater. Reverse engineering regulatory algorithms will be child’s play for the next generation of AI.
Bank fraud and Dieselgate: how do we design regulations that are harder to cheat?It's easy. Have you tried a board with a nail in it?
Regulations you say? HAHAHAHAHAHAHA!
The testing criteria should be stated to sufficient (lack of) detail, and indistinguishable from real world conditions, such that any defeat algorithm will have a hard time telling the difference between a test lab and real world conditions.
Also the test creator should consider that the test submissions could also be data mining for determining the nature of the tests. Randomization of test type and length should be sufficient to thwart these data mining efforts. Ultimately this test vehicle should be taken out of the lab and (with test instrumentation on board) be subjected to some real world conditions. Final spot checks should also be replicated on random production vehicles in the marketplace.
Wanderlei Silva was barred from ALL competition in his field of endeavor for avoiding one urine test.
Put the people in charge of USADA and WADA in charge of penalties for the motherfuckers.
Or just kill them outright.
There is no single ‘way’ to identify cheaters. This is an ongoing war and the best we can do is adjust to each elevation in the contest.
It might help if the testers/regulators were, themselves, able to think like cheaters rather than rational test engineers. Perhaps we need to bring in magicians; they are among the most critical thinkers around, mainly because they understand the psychology and reality of the cheat.
I’d also suggest that if more corporate heads wound up in jail, they might come to a new understanding of the cost;benefit calculation.
I think Volkswagen should have been fined to the full extent of the law, banned for selling cars in the US indefinitely, and its plants sold off to its competitors at fire sale prices. Good bye VW, Audi, and Porsche - see you in the distant future. maybe.
Here’s an idea: regulators can be vaguer by applying Monte Carlo methods. It’s probably not even a new idea.
All corporate fines and taxes should be solely on the investor’ profit. How it is structured now all fines and taxes are passed on to the consumer and employees.
Investors might use something beside profit margin to invest, and steer the corporation in the right direction.
How do we design regulations that are harder to cheat?
… is a question never to be asked by 45’s administration.
the penalty for being caught cheating should be having the company broken up into pieces so small they could neither offer effective bribes to regulators nor scare courts with the fallout if they were meaningfully punished.
Corporate death penalty. It’s not a bad idea. The main problem with it is that its likely to let the leadership get away with destroying the company while putting the burden of punishment on the investors. It needs to be done in tandem with real criminal penalties in place for the management who either push their employees to cheat or look the other way when they do it. And ignorance is no excuse - if the company doesn’t have checks in place for senior management to find out that cheating is going on that should lead to the same penalty as ones that do but ignore those checks.
Without effective punishments for the pirates who swoop in, play tricks to prop up the stock price, steal investor money, and then leave via a golden parachute even breaking the company up into tiny pieces won’t put a dent into this kind of cheating. Because why should they care? They got their money. The ex-CEO of VW got something like 9 million Euros when he resigned. What’s to stop the next pirate at the head of, say, Nissan from doing the same thing even if VW is broken up into a hundred different companies? Nothing - because pirate CEOs for hire don’t care about the long term health of the companies that they pillage, just the amount of booty they can stuff into their bank accounts before they get caught.
I think the punishing of the shareholders is justified in this case, not to go any easier on the executives, but if the shareholders do not suffer for the actions of the management team they put in place, then nothing will change.
On the other hand if Shareholders start to routinely loose the full value of their investment to a Corporate death sentence, then I think corporate oversight might take a step towards improvement.
I support that idea of vigorous criminal prosecution of people in the C-suite. Too big to fail is a problem because of collateral damage. But putting people in jail for long periods of time is very narrowly targeted. Furthermore, no matter how rich you are, you’ve only got a limited number of years on this planet. So taking away those years hurts whereas fines (aka bribes) do not.
It came to my attention recently, that many other vehicle manufacturers also cheat on emissions.
Their problem is in complying with cold start emission requirements. It’s just hard to make a cold engine burn cleanly.
So what do they do?
They add a pump, to pump fresh air into the exhaust so that the emissions out of the tail pipe are diluted and the parts per million of emissions as tested at the tail pipe look good.
It does not reduce emissions at all.
It just fakes it for the test, adds cost and complexity, and since it requires power to do the pumping, will actually produce slightly more emissions because the engine has to do slightly more work.
That’ll show those employees, dealers, and supplier chains!
On emissions: Colorado does random drive-by emissions testing. They set up a van with a sampler and camera, narrow traffic to one lane and capture everyone who happens to be driving by. I encounter one of the drive-bys about 4 times a year. Real world conditions, in a location where cold start is statistically unlikely, with multiple samples over the course of a year. Anecdotally, I know multiple people who drove diesel VWs and Audis who got emission test required on their registration renewals, because they failed the drive-bys. The standards for the test are completely transparent; only the time and place are undisclosed.
On banking: I read The Girl With The Dragon Tattoo around the time of the crash, and what stuck with me from that book was the idea that the financial press should not be cheerleaders for the financial industry, and financial regulators should not be advocates for the industry. Both should be adversarial, skeptical relationships, The SEC’s revolving door is a known bug; if an SEC auditor knows they will be job-seeking in the industry in 3-5 years, it is in the auditor’s personal best interest to go easy on those they’re regulating. If we can regulate that aspect, and figure out a way to do the stress tests at random times with no warning, we might be able to maintain both the transparency of the testing method and the usefulness of the test.
It will show them. It will show them that the cars no longer being made and sold by VW will be made and sold by their competitors. And those employees will go to work for those competitors, and those suppliers will supply parts for the new extra cars those competitors are making.
Service people and salespeople from the dealers will similarly find work at their competitors who are now selling all the cars that VW used to sell.
The Dealership owners? They are bought out by VW as part of the punishment.
Everybody is whole, VW is gone. Thank you.
Yes, exactly. If the only maxim followed is to create value for the shareholders then you have to make cheating disastrous for the shareholders. That has a lot of negative consequences, but those are the consequences of choosing a dog-eat-dog system. It’s just that currently investors are protected from those consequences.
This is what has to happen. Why even bother talking about how to catch cheats when the penalty for cheating is a fine that is less than the amount of money you made by cheating?