= 27d567bc2d48d9f40e9ccc20ea370b15)
That still seems to leave room for (1) one or more broad cycles where most of the watchdog industry grows rich off lies & bias before imploding (as happens with bubbles) or (2) a steadily growing industry with a continually replenished proportion of fraudulent watchdogs that pop into existence, make money from lies, then get swooped on (as happens with consumer fraud).
Where Are We the Weakest?
As rationalists, we should be able to consistently and accurately make predictions that enable us to act effectively.
As humans, we don't. At least not perfectly.
We need to improve. Many of us have, or at least believe we have. However, it's a notably hacked improvement. PredictionBook is an excellent source of feedback on how well we're doing, but there's more detailed information that isn't easily available that I think could be incredibly useful. Questions I would like to see answered are:
- What kinds of predictions are we the least successful at predicting? (weakest calibration, smallest accuracy)
- What kinds of predictions have the most low-hanging fruit? What's the easiest to improve on right now?
- What kinds of predictions are the most useful to us? (accurately predicting a close friend's behavior>predicting obscure political decision)
- Where aren't we making quantitative predictions? Where does our behavior involve predictions that are underrepresented on PredictionBook?
Before we are able to improve as a community, we need to know where to improve. I'd love to hear suggestions on how to answer these questions in the comments.
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
It seems as if these are generic enough problems with enough past incidences to assume studies have already occurred regarding them. Do you happen to know what the existing solution to these problems are and how effective they are?
or alternatively, how harmful the problems are, and wether an imposed solution tends to be cost effective or not.
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
Is this correct? Have you communicated the correct ideas to me?
From my experience, yes.
And, In what ways have business credentialing systems handled this kind of demand historically, and are similar methods applicable here?
In my experience (which is by no means exhaustive), the "recommended" solution has been to ensure that the data entry and transmission processes are performed by humans, rather than fully automated, and that no automated data integrity checks are performed.
Also in my experience, this 'solution' does far more harm to the system than the nepotism itself. :(
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
Ok. So the problem is that designing systems with accessible back-doors for nepotism is a larger hit to the effectiveness of the industry than if there were simply blatant nepotism itself, and systems weren't compromised by third parties.
Is this correct? In your (admittedly non-exauhstive) experience, in what specific ways is the industry harmed?
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
I don't see how nepotism effects the system too horribly. Businesses are still entirely allowed to hire close friends and family despite less credentials, and any other nepotistic behavior that's counter to the purpose of profit.
Well, I was speaking from the perspective of someone who has actual experience in writing business credentialing systems. The effect is small enough to come out in the wash most of the time, but is important enough to require implementation (or at the very least, to prevent robustness in the implementation of goal 1.) The problem is that profit-maximization is ultimately less important than status-maintenance, in many cases - but the environment requires that profit-maximization appears to be the primary motive.
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
Ok, I think I understand what your saying.
The system needs to be designed in such a way to allow covert nepotism, as businesses desire the ability to behave nepotistical without appearing to be, and this demand is important and consistent enough for the market solution to inevitably include it. And you are also saying that nepotism isn't too harmful, and the system we'd attempt to implement should include mechanisms for nepotism.
(God I sometimes hate human behavior. ARGGH)
Is this correct? Have you communicated the correct ideas to me?
And, In what ways have business credentialing systems handled this kind of demand historically, and are similar methods applicable here?
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
Businesses wan't an accurate idea of who is employable.
This is only partially true. Businesses want two things:
- An accurate metric of who is employable
- The ability to co-opt that metric for specific people, in order to promo te nepotism without producing the appearance of nepotism.
'2' is a rare exception, but it is a VITALLY important one - important enough that any system you design will be doomed to failure if it fails to provide for case 2, or makes its process of providing for case 2 too obvious to outsiders.
EDIT: Okay. I'm going to go ahead and take whatever further karma hit I have to to point this out. I lost FOURTY EIGHT karma over the course of five minutes, after posting this. Seriously, LW. Fix your goddamn karma system.
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
I don't see how nepotism effects the system too horribly. Businesses are still entirely allowed to hire close friends and family despite less credentials, and any other nepotistic behavior that's counter to the purpose of profit.
The Watchdogs, however, can't just hand out certification willy nilly, because they need a universal metric for all competing businesses in a field.
If Business A makes a habit of bribing Watchdog A so Mr. CEO's sons can have the same certification as actual students (for the purpose of hiding nepotism), then Watchdog B's nepotism free selection methods will locate higher quality employees more consistently, and garner costumers from the majority of companies that are primarily interested in actual profit.
In response to
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
Who knows what amazing research teams might be borne of the intellectual equivalent of OK!Cupid.
This sounds like an awesome idea. Does anyone know if such a thing already exists? If not, I would be willing to commit $500 to someone who wants to make it (assuming they already have the appropriate skill set, of course).
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
I would like to strongly voice my interest in such a service myself. I don't have spare capital to give, but I want this to happen.
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
So the reason employers wouldn't make the mistake with testers that banks made with moneylenders is... that there isn't a secondary market of diplomas/graduates? That educational entities with a vested interest in getting evaluated well would be a much smaller part of the market of employers than banks with a vested interest in selling bad debt were in respect to the market of debt buyers?
Am I getting this right, or am I just making a spectacle of my ignorance?
In response to
comment
by
on
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
I would think business wouldn't make the mistake of employing biased watchdogs. Let's look at the main incentives:
Businesses wan't an accurate idea of who is employable.
Schools of course want to appear to produce employable students.
Watchdogs want money.
Now, unless I'm mistaken, being bribed by schools to put them in a favorable light is a temporary bubble, liable to pop when businesses discover that they've been lied to. Watchdogs' main employers would be businesses searching for employees, as far as I'm willing to predict, and failing to successfully out-predict competing watchdogs will result in a watchdog's failure.
Fraud will happen. Humans tend to lie for short term gains. However, it's an unsustainable business model in the long run. Fear it greatly as this market is forming, but as a trend it shouldn't last.
In response to
A Gamification Of Education: a modest proposal based on the Universal Decimal Classification and RPG skill trees
Absolutely love this idea, just one little comment on the watchdogs.
It occurs to me now that, if one wished to be really nitpicky about who watches the watchmen, I suspect that there would be institutions testing the reliability of those meta-institutions, and so on and so forth... When does it stop? How to avoid vested interests and little cheats and manipulations pulling an academic equivalent of the AAA certification of sub-prime junk debt in 2008?
Why do the watchdogs exist? Because business produce demand for accurate measurements of employability, and are willing to pay for a service that offers that information. I'm sure if a significant number of watchdog companies exist a business will need to employ a meta-watchdog to determine which watchdog to employ, but it does eventually terminate at the employer themselves, who has to personally be the watchdog for which meta-watchdog company to employ.
And that information is easily available (albeit after some lag). Which company reliably recommends the tests that actually produce better employees?
How big is the chain of information when you're deciding which restaurant to eat at? That's perfectly analogous scenario, as far as I can tell.
In response to
July 2013 Media Thread
Music Thread
For the EDM fans out there, some semi-pony themed dubstep/progressive house/electro music: http://www.youtube.com/channel/UCj5ceb5aLKALh-g6vhAcpxw
View more: Next
= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
= f037147d6e6c911a85753b9abdedda8d){kind=logo}
Ok. I naively thought I could trip up this system by altering the probability of being a simulation or the payout, but I can't.
SCDT successfully 1 boxes on any box 2 payout larger than $1000. SCDT successfully 1 boxes even if a single simulation is used to predict any number of actual Alices. (The scenario I worked through involved 10,000 duplicate Alices being predicted by a single Alice simulation.)
I'm thoroughly impressed by this decision theory.