Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror

Comment Re:executive malfeasance bonus repatriation (Score 2) 77

Furthermore, in game theoretic terms, the upside is all the unethical scams combined, and the downside is only those which you eventually get caught (and these with a fat net-present-value depreciation term).

This is why the controlling stakeholders set up bonus conditions that the senior executive ranks can game to their personal advantage.

The fines on the ones you don't get away with simply aren't large enough (historically) to deter a general ethos of catch while catch can.

It's an extremely tough problem in mechanism design how to incentivize executives to cut the legal corners where the aggregate ROI is positive, without incentivizing the executives to cut the rare, extremely dangerous legal corner where the ROI is a nightmare on wheels (Volkswagon might be in this later camp, although set against that, the threat to their diesel consumer franchise had they not cheated was possibly existential).

The stakeholders have a severe oversight problem, because you want plausible deniability on the upside scam, which makes it hard to formally poke into what the management team is doing (leaving actual breadcrumbs behind) when you suspect they might be playing more than a little over the edge.

Breaking the law just right: priceless.

That's why they make the big bucks.

Comment executive malfeasance bonus repatriation (Score 1) 77

Yeah, it was pretty dumb of T-Mobile to do this and getting caught. They knew the rules in the US and that punitive damages can be quite high.

They (the actual humans who work there) also know that annual bonuses are profit driven, and that those bonuses are rarely clawed back when it turns out those profits were delivered along with a ticking time bomb hidden under the floor boards.

Executive bonuses can be fabulously remunerative. Gaming internal performance metrics is also open season.

These slap-on-the-corporate-wrist punishments would work so much better if they mandated executive malfeasance bonus repatriation.

Comment the long tail as Disney gremlin milkshake (Score 1) 93

I have a list pushing 700 titles of feature films/documentaries that I've viewed over the years (which I started to fill out retrospectively about six years ago, as it became harder and harder to sift the seen from the unseen).

And I have another list of 400 titles with decent critical regard that I have yet to see (few unseen epics remain, but the list is not terrible).

My local library has about 80% titles on DVD and my local video store (20,000 item catalogue) has another 19.5% (long may it thrive). Total expense: about $30 per year, plus those annoying trips to the library, where I also borrow books, at least weekly.

Our total television consumption is the roughly 50 movies we rent/borrow each year. It has no other hookup.

The whole point of the long tail (for me, anyway) was that finally I could make my own list primary to the choice process (instead of the shifting sands of offered catalogues). For my purposes, the long tail has always been the only killer media app.

Any prospect of the average, unthinking sod toppling into my camp makes Disney very unhappy. And so they are busy feeding long-tail gremlins into the wood chipper, out behind Burbank's aged Annette Funicello building.

The mouse never forgets it's original business model.

Comment Re:Eneloop is the way to go (Score 1) 210

If you need more than 1.2V out of your alkaline battery, you will chew threw batteries pretty quickly. A typical discharge curve has about 40% capacity left when the alkaline hits 1.2V. Generally you don't consider an alkaline depleted until it's at 0.9V.

Nothing profound here, just a straw poll of my own records, since the subject came up.

I usually measure my "dead" batteries (no load, but usually immediately after attempting to draw power from the device, which does make a difference).

One of my kitchen scales spits batteries out at 950–1050 mV, the other scale goes into calibration wobble (a few 100 mg, but annoying) south of 1300 mV (three batteries).

Even so, I get about the same year from each scale, on regular daily use, given good quality alkalines (the precise scale is faster and better, but only goes up to 600 g). Both scales are nuts-on my 200.0 g calibration weigh, year after year, to within one ULP.

My quad kitchen timer spits batteries out at 1150–1250 mV. This is LCD-based with no backlight (but LED activity lights), and usually lasts 18 months. Every time I start any large cooking project, I set a 6-hour countdown timer, to track absolute elapsed time. It doesn't always run to zero, but it still racks up a lot of hours. The idea with this timer is that you can time each hob separately, but that doesn't really translate in practice in a busy kitchen. What I love about this timer is that it keeps counting down into negative numbers when I blow it off, to deal with something else more urgent. None of my other timers measure my blow-off delay (which can be important in tracking total cook time).

My laser-infrared kitchen thermometer rates itself for 3300 15-second activations on two AAA batteries, and I think that's about right, but I didn't get around to measuring the failure voltage on the last battery replacement cycle. I don't think it would like 1.2 volt batteries, but you never know.

The only devices I have that would run all the way down to 900 mV would be cheap LED flashlights.

I sent a third kitchen scale to my wife's (outdoor) barn to measure horse nutrients. She brought it home once after several years in "use" for a battery service and both batteries were still at 1550 mV. So I cleaned the contacts and sent it back.

The problem with NiMH batteries is that almost everything I have that sucks any serious juice bonks out too close to the 1200 mV range. I have a Sony voice recorder that I use almost daily (actually designed for NiMH as well as alkaline), and I've actually had to replace several sets of NiMH batteries that ran out of charge cycles. Hmm, I never thought about this much before, but it explains why the top charge bar disappears with the NiMH batteries about 30 minutes into their first use after a fresh recharge.

We once had a Sony Walkman that I once used to capture my wife's favourite mix tapes onto the iPod. I guess country music isn't impacted much by capstan wobble. Anyway, she was completely happy with this, even though I think the beat faded at the end of each battery pair (actually, I think it just became less capable of compensating for variation in tape tension, because it was surely regulated to some degree).

Comment universal magic bullet secret powder (Score 1) 270

Every magic bullet with any claim to success devolves onto the same success factor: hiring better people, under progressive management.

This why many magic bullets enjoy their 15-minutes of fame, but few magic bullets scale broadly beyond those happy beginnings.

The underlying scarcity is probably incentive alignment: it usually takes a rare syzygy to gather together the best groups.

A Solvay Conference is not in the pyramids any old century.

The leading figures were Albert Einstein and Niels Bohr.

17 of the 29 attendees were or became Nobel Prize winners, including Marie Curie, who alone among them, had won Nobel Prizes in two separate scientific disciplines.

This conference was also the culmination of the struggle between Einstein and the scientific realists, who wanted strict rules of scientific method as laid out by Charles Peirce and Karl Popper, versus Bohr and the instrumentalists, who wanted looser rules based on outcomes.

Starting at this point, the instrumentalists won, instrumentalism having been seen as the norm ever since, although the debate has been actively continued by the likes of Alan Musgrave.

Even then, there's always some slack-assed Jules-Emile Verschaffelt in the room, to screw things up.

Comment Re:It's time to user smaller specific social media (Score 1) 291

They answer "yes/no" if they say yes, we use FB later to exchange ... when one of us remembers and thinks it is "worth it", if "no" we exchange the eMail address. Chances are, I lose it on the way home ... or forget to mail.

A beautiful world it would be if all women not intending to conceive were on the pill all the time.

Even better: no possible downside.

Well, apart from the odd course of antibiotics after the fact (supposing all those overused antibiotics still work ten years from now ...)

But, whatever.

No worries.

Comment turgidity on wheels (Score 1) 64

Both of the "we suggest" paragraphs are coma inducing, so you get past that and what remains? This:

At the same time, the potential commercial rewards from mastering this mode of research are likely to usher in a period of racing, driven by powerful incentives for individual companies to acquire and control critical large datasets and application-specific algorithms.

Strunkian synopsis: Imminent ruthless-genius land-grab race to the bottom.

Then the rest of the words could have been devoted to explaining precisely what they mean by "application-specific algorithms", because specificity roams wide.

Are we talking 120 or twelve thousand?

Comment Re:Inventor of the world wide web ... Oh please! (Score 1) 180

Hyperlinking already existed by the time Tim Berners-Lee re-invented for it the THIRD time.

I'm almost surprised you didn't include Ada Lovelace.

Part of the invention package is planting a novel idea into a viable context. Farmers figured out this trick 10,000 years ago. It's called soil. And guess what? Half of every tree is immune to lovers' penknives.

Context, the Rodney Dangerfield of being in the right place, at the right time.

Comment the economy of "complete thoughts" (Score 1) 16

... still has a long way to go when it comes to connecting ideas and answering questions with complete thoughts.

Customer: [twenty-word utterance of word salad]

Google Books: Please express a complete thought.

Customer: [doubles down on word salad]

Google Books: Not even close.

Customer: Find what I want, or I'm going to bleeping tear you a new one!

Google Books: Yes, that does qualify as a complete thought. Unfortunately, it's not searchable.

Customer: You are so getting it, as soon as I find my baseball bat.

Google Books: While you rummage, here's some interesting trivia for your consideration: the strange thing about the Lernaean Hydra is that nobody ever checked the other end.

Customer: What's that supposed to mean?

Google Books: If every simple tube has two natural ends, your adrenal imprecations are only of marginal concern to this particular terminal endpoint.

Comment Re:Money-Grubbing Sociopaths (Score 1) 368

Since when is $12.5 billion in initial sales and $4 billion a year not a profit?

When it's just one slot on the roulette wheel, and most of the other slots involve risk bleeding out the wazoo (development failure, competitive risk, lawsuits, monstrous capital-investment cycles involving Nervous Nelly moneybags).

No-one writes position papers on risk-free, certain-profit outcomes, because they're too busy scraping $100 bills off the sidewalk.

Let's briefly consider the curious case of Michael Burry.

Though he suffered an investor revolt (where some investors worried the logic was wrong and withdrew their investment in Scion Capital's hedge fund) before his predictions came true, Burry earned a personal profit of $100 million and a profit for his remaining investors of more than $700 million.

Here's the story of one of the heroes of 'The Big Short'

At the same time, Burry began to tell his investors of the enormous risks to the system. His investors were mostly institutions that did not want to hear his theory. Their other investments were all built upon the concept of a sound system with no subprime mortgage risk. Investors began to get nervous and demand their money back.

Unfortunately, it was too late as Burry had already gotten into several long-term, illiquid bets against the market using derivatives to bet the price of mortgages would fall. If he got out of the trades, he would suffer a huge loss — so Burry simply refused the investors' requests.

The film depicts lawsuits against Burry originating from the same people who later walked away with the $700 million, who could have easily scuppered the whole position, if Burry hadn't been such a prudent ass.

There are two main drivers of (apparently) obscene profit: rent and risk.

Often rent is the solution to risk.

That's basically the founding model of the modern corporation. Bankruptcy functions as a hard cap (in most circumstances) on investor liability. Many corporations have engaged in risky bets with enormous upsides and downsides, from which they later walked away (just the failed half of their portfolio), lobbing the majority of the downside to the government (aka you and me), which then faces a billion-dollar cleanup project.

The built-in rent of limited liability makes many business models viable that would not otherwise get off the ground.

It's actually a legitimate question whether a little bit structural rent greases the skids and makes the world go round.

Rational ignorance

Rational ignorance is refraining from acquiring knowledge when the cost of educating oneself on an issue exceeds the potential benefit that the knowledge would provide.

Milton Friedman was big on this (he used to go after the sugar subsidy on this foundation):

Sugar Subsidies, America's Least Efficient Corporate Welfare Program

Washington is once again massively screwing up the American sugar market. Because American farmers cannot compete with foreign sugar growers, the federal government has maintained an array of sugar import quotas and/or tariffs for most of the last 200 years.

As a game-theoretic principle, rational apathy has wide application.

For one, it's really hard to see how a purely libertarian free market deals with the chlorofluorocarbon problem, where the downside is so diffuse (and distal), that no-one at all has a local incentive to begin the (costly) process to self-organize their interests—entirely mediated though a viral-groundswell freedom to contract—to protect themselves from a dawn to dust SPF-90 reliance in a UV-enhanced future.

In this example, regulation is the market-efficient solution.

Any way you slice it, risk and rent make for strange bedfellows, and no number should shock on face value, without first applying a serious critical reflex.

Comment Re:If that claim is true,.. (Score 5, Insightful) 172

There is actually a way to make sense of this: a lot of people deleting mostly dormant accounts, with hardly any content posted, having almost no discernable impact on Facebook's daily churn.

Basically, it would be that group of people who only filled their Vicodan Rx once, and never actively sought a renewal to begin with, all suddenly flushing their nearly empty pill bottles, after a news report goes out that the smell of Vicodan pills is a randy Bugblatter aphrodisiac (but no-one else).

Comment Re:If that claim is true,.. (Score 2) 172

This reported number doesn't pass my smell test, either.

With such a precipitous drop in users, you'd probably hear the entire fabric of the universe groaning and throwing off glowing metallic divots as it passed through some kind of nearly impenetrable Wrong Stuff barrier.

If true, this story would already be the Mount Krakatoa of the social media era.

Maxwell Smart: Would you believe "1 in 10 are thinking about maybe deleting their account"?

Comment porcupine entropy (Score 0) 139

These fixes that achieve practical "randomness" actually make the RNG LESS RANDOM, but more secure for some models!

I guess you never played hangman with a blood lust. Adversarial randomness, it's a thing. Eventually you reach a game-theoretic equilibria. The equilibria will never assign a probability of zero to any password.

Your underlying mental model here is that this is a multiplayer game, with a large group of guppies, a smaller group of porcupines, and some community of crackers.

New rule: guppies don't understand porcupines.

New rule: guppies barely understand crackers.

So the guppies will end up at a game-theoretic solution which is far from an optimal strategy.

New rule: the crackers don't know the guppies from the porcupines when starting to crack a new password.

So the crackers will adopt a hybrid strategy to maximize crack rate based the population of guppies and the population of porcupines. No matter what strategy the crackers adopt, the guppies basically amount to a fixed point. This also means that the crackers will prioritize exploration of the guppy ghetto ("God", "password", "12345678", etc.) regardless of how the porcupines behave.

From the cracker perspective all the non-randomness derives from the guppy population. Asymptotically, as the guppy population shrinks, the porcupines will adopt a uniform distribution over the entire password space.

Essentially, porcupines avoiding "password" only looks less random if you advertise that you're a porcupine to the cracker population. If they really take you seriously, they wouldn't bother to check "password" early (advertised porcupines would be presumed to use a fully random password).

But it costs nearly nothing to check your bluff by running the list of the one trillion most common passwords, and this whole Dr Strangelove "tell them" strategy presumes 100% of the crackers actually notice your "I'm a porcupine" disclosure.

If any of the crackers fail to notice (and to automatically take your disclosure seriously), you don't want to be using "password" at all, ever.

Why is the game-theoretic embedding so intricate?

Because the first trillion (or first trillion trillion) most-common passwords are numerically insignificant in a 60-bit password space.

So avoiding "password" & co. doesn't dent your entropy in any digit anyone would ever bother to write down, and the whole story I've just told is asymptotic to a distinction without a difference.

In summary: no, the entropy goes up when porcupines filter out non-starters, as measured from the appropriate game-theoretic node (crackers who are locked into the strategy of not distinguishing porcupines from guppies before the attack begins).

Here's a second-order asymptote to wrap this up: if the cracker really, really believes that you are a maximal porcupine (you're a fully-upgraded positronic borg descendant of Colin Percival or Moxie Marlinspike, with the factory anti-tamper sticker on your emotion chip in pristine nano-crystalline condition) then the cracker doesn't test a single password at all with a classical computer: it would be attojoule wasted to no conceivable economic upside.

For the cracker to be stuck with a classical computer, time travel would probably be involved, and the cracker would be stuck in some prehistoric nowhereville using the mouse, instead of talking to it, or merely holding it to his forehead.

But hey, it could happen, so a true porcupine needs to be prepared.

Slashdot Top Deals

System restarting, wait...

Working...