These are a couple of posts I made 2 and 3 years ago and I’ll be darned if both of them might have a little value today with complexities and everyone looking for the “magical” algorithms as a one shot solution. I don’t call myself a predictor at all but like to go back and see if any speculation I did comes around.
Sure algorithms are great and create a lot of neat things in science
and otherwise but when you get down to the day to day grind of wanting a one stop solution especially with analytics and business use, it’s just not there. We hear CEOs putting out press releases and other media that “we can save a trillion dollars” in healthcare which is simply not true. We can save money but it sounds like someone has meddled in some extreme non linear analytics that has not been verified. As Charlie Siefe says on models, the average consumer reaction and thoughts are something like this “it has a square root in it, I have read about it several times this week so it must be good”. CEOs do that too.
Here’s the original post from around 3 years ago when the Madoff case broke and shoot even he was smart enough to hire IT folks to create his “fake” front which later was exposed by his former IT folks. Harry Markopoulos went nuts over this and for good reason as there’s a lot of smart people but we don’t listen to them.
“Department of Algorithms – Do We Need One of These to Regulate Upcoming Laws?
A Case of “Dirty” Algorithms – 2 Madoff Computer Administrators are Indicted – Illegal Coding and Networking for Big Profits
I’ll bring this one up again about the PLOS One study with people fearing math and feeling pain. Of course I am using a little satire here, but gee based on just my own experience with talking to people, watch them run for the hills when discussing even something very simple. A few years ago when I was doing some software work for a hospital I had a CEO client that would do that when I walked in with my tablet. “Oh my gosh she’s in here with that thing again”…you could see the look and now look where we are, everybody does it:) You can see it’s very easy to get Algo Duped today as this stuff gets very complex. Algo Duping 101 with a series of videos can be seen here or just watch the 5 videos on the left hand side of this blog. Let’s start shoving every member of Congress in an MRI and see if this fear exists with the folks we elected:)
“Algo Duping” – PLOS One Journal Publication Explains Why The Fear of Math Plays a Big Role As One Underlying Reason We All Get Duped And Those Who Don’t Fear Math Take All the Money, Gradually, Using “Mathematical Formulas & Algorithms”
Fact is nobody is around checking any math and coding here as there aren’t that many that can or they are around and people don’t use them or want to pay for it. Just run the code if it makes money. Last week with Ticketmaster and the inauguration ball, perfect example. They sold the tickets by mistake ahead of their advertised time of when they were to go on sale. Did they do anything to make it right..nope. Those who went to the website early got tickets and their faithful clients who waited for the day of the initial sale were screwed. Lot of companies doing business this way today and consumers get screwed as they don’t want to go back and mess with their data…hmmm…a place for a Department of Algorithms here?
Presidential Inaugural Ball–Ticketmaster Algorithms Blow It–Tickets on Sale Before Announced Sales Date– Any Companies, Markets, Etc. Know What They Are Doing Anymore With Models and Algorithms? Attack of the Killer Algorithms Chapter 51
Actually we need a balance between intangibles and tangibles to create jobs here in the US so more of the technology can be used in manufacturing and not just flat out mining and selling data for billions in profits. That’s why companies don’t expand as it’s much easier to take the easy way out to mine or take data they have in house and sell it. Wall Street totally exaggerated the value of algorithms like with Facebook and those algorithms bit back with the IPO opening day with Nasdaq and the dollars are still not settled.
“Devaluate the Algorithm” And “Tax the Data Sellers”–A Cure for Both Healthcare and an Economy Based Heavily on Intangibles–We’ve Lost Our Balance
I’m beginning to feel like I said something that has some value here with accuracy and holding modelers and programmers accountable:) Here’s a sample post on the complex algorithm at the link below used by Medicare for hospital compensation and even they can’t work with it in house and are fixing the computational algorithms:)
The Complicated Algorithms Used by Medicare to Establish Hospital Compensation
Moving on a bit I do a lot of reading every day and see this software, that app and this website all have the ultimate answers to what ever ails you it seems. Having written software before I have a bit up on someone who had not and there too are much better programmers out there than me too, but so much I see is repetitive and some has just little or no use and developers if it is a consumer product that don’t even use their own stuff. I read not too long ago from a developer that said the same thing I just said here, use your own software if it is a consumer product:) I see data bases related that don’t even belong together, time to make algorithmic bucks is what it is.
I don’t care if people want to crunch numbers with different data bases as you can do that SQL work all day long; however if it doesn't have value and is created to make a buck only, then stuff it and stop writing code to just make a buck. FICO is one huge example of this and even the layman can see it with using credit score and other non credible data queried with it to sell analytics services to pharma and insurance companies. I say it’s a far stretch here as if they buy it and query with gosh knows what, you end up with real junk that hurts consumers that the average employee is told to look at the screen and use the data presented to them for their decision making. If you have not seen the far stretch of what FICO is doing for a buck, read it by all means. They are really out there on this one and should be admonished and their code should be checked and sources revealed by all means. A place for that Department of Algorithms….what are they doing to to now with potentially mixing this with revenue cycle business?
FICO Buys CR Software And Gains Access to Revenue Cycle Software Business in Healthcare
FICO Analytics Press Release Marketing Credit Scoring Algorithms to Predict Medication Adherence–Update (Opinion)
So what we have here is analytics used out of context and watch out as this insane analytics to make a buck hurts all of us. I did a beg detailed post on that as well, again nobody checking any math, models or logic, just the company selling algorithms….Department of Algorithms…
the information on the website is set up to “dupe” you with thinking there’s value here, not. So even if the model has a square root in it, no value, this is one of those with FICO. There’s a great video at the link below with big companies on a panel discussing “value” and how to find it.
Big Data/Analytics If Used Out of Context and Without True Values Stand To Be A Huge Discriminatory Practice Against Consumers–More Honest Data Scientists Needed to Formulate Accuracy/Value To Keep Algo Duping For Profit Out of the Game
This is a good podcast radio show and listen to it on how people lie but numbers do not and then watch his video on the left hand side of this blog and what I say here will begin to make more sense.
“Numbers Don’t Lie, But People Do”–Radio Interview from Charles Siefe–Journalists Take Note, He Addresses How Marketing And Bogus Statistics Are Sources of Problems That Mislead the Public & Government
Sure we have Big Data and machine learning in the future but again it’s the value of this work that needs to be shown as well as accuracy in computation and not code for just making a buck so be careful when you read some of what is out there today and be aware of Algo Duping for cash. Here’s a video from Cathy O’neil who worked as a Quant on Wall Street for a Hedge Fund and this was done by PBS Frontline. She’s far advanced over my background for sure but she explains the same things I have been talking about with fantasy formulas and how models and and algorithms are used against you.
Watch The FRONTLINE Interview: Cathy O'Neil on PBS. See more from FRONTLINE.
Now for a little fun…with all this going on and being promoted for value whether it is there or not, how are we going to deal with some of this…are people addicted? Sure they are in more ways than one. 
Good scientific and other good business intelligence algorithms will continue because it makes us smarter but we will have this function of trying to determine what’s a dupe and what’s real…so here’s the next 12 step program for those that think an algorithm is a cure all without the human touch and for those that create the nonsense that clutters the good stuff:) I made this post back in August of 2010 so ahead of my time maybe? You tell me.
Data Addiction and Abuse –The Up and Coming Next 12 Step Program Is On the Horizon–Side Effects Include Lack Of Data Quality, Integrity And Spasmodic Algorithms
We end up with something along this line…and those spasmodic algorithms are causing everyone a bit grief with lack of integrity.
“In God we trust; all others must bring data”
Nobody has done squat to help engage the consumer so far as they are so wrapped up in analytics, the patient is still left out in the cold and heck we don’t even see any leaders using any technology themselves. So with these two old posts…was I on track here? 
It’s not that I’m so smart it’s just that people who work in the technology field get there before you do as a consumer so sometime we know a thing or two. Should we work on getting an agency started who can help verify accuracy of some of the stuff tossed our way today and should we begin looking for those that are developing a crippling addiction to cure one algorithm with another? Again these were topics from 2009 and 2010…and they still seem to have some merit for conversation you think? BD
No comments:
Post a Comment