So, do you wonder why we can’t create more new jobs? Well when one can hire a few technologists and mine data and create analytics by the barrel (which half will end up being useless) why would a company expand with tangibles when a few lines of code and some algorithms can made even more money that manufacturing a product? There you have it, data selling. The data selling business has reached an all time frenzy and for lack of better words has become an epidemic in the US and maybe the world for that matter. When I say epidemic I mean just that. Value is NOT created with every SQL statement or query that someone creates with data bases, learn that fact early and be aware. If you read the web we are now seeing articles talking about “fake data scientists”. Would there be articles about this topic if there were not fake ones running around out there, of course not. There are. As a side point here I made a satire post about 3 years ago about data selling and I’ll be darned if it is not almost true about needing a 12 step program of it’s own. I guess you could call this yet one more example of satire becoming reality. This post was a hit around Forbes and a few other major publications:)
Does the US Government Need a Department of Modeling and Algorithms–Is Data Addiction and Abuse the Next Up and Coming 12 Step Program, Some Classic Posts & Topics Revisited
The entire data selling business was originally set up to help people market you better with relative products and it might have not been too bad if it had just remained as such but it has not. There are all kinds of analytics that claim to provide you with so many unlocked secrets of predicting the future. Granted there’s value in crunching numbers for sure but not when you take models down to an individual level where nobody has created or questioned a “P” value, in other words did they write code just to make money? Yes that is done out there with matching obscure data sources that “they “claim can be queried for tons of information relative to predictive behaviors.
If some of this is used in context certainly there is some good trending information that offers direction, but not the way it’s being done and over sold today. The only folks who are really lining up to use most devices today are those who “have” to due to a medical condition where the numbers and stats report back to a web service of some sort, otherwise there’s very little interest. Sure we have all the consumer apps and even those with the best interests at heart get bored and stop using many of them, just ask them. Happens all the time. Again knowing that someone is selling that data whether it be a profile or other more detailed information is enough to stop anyone from wanting to use one.
If you don’t think dirty code and models are being created just to make money, then check out this next link to where the folks who have the talent to create models and algorithms to do so are creating their own code of ethics, which includes a couple sections about having the liberty of reporting those requesting numbers and formulas beyond what is ethical and legal to law enforcement. The only problem with this is having folks in law enforcement that are smart enough to see and understand what is being done. If you watched the PBS special on “Too Big to Jail” (video at the link) then you can hear in the first 5 minutes with DOJ prosecutor Manny talking that he absolutely had zero confidence and that this type of investigation was way over his head. By the way he just got a new job after resigning from the DOJ making over $4 million a year. Sure the data scientists want jobs to help the betterment of man kind and create efficiencies but want the 200 pound “greed gorilla” removed from the process so things are done correctly.
Big Data–The Data Science Code of Ethics-Designed By Those Who Create Models - Don’t Fall Victim To Write Fictitious Code and Models Just to Make Money With Clients Demanding Such
You have social networks selling data about you right and left and some apps want you to include social networks so it can all be linked. More data to sell for sure. You should read this post on when machine learning makes errors as the social networks all want your “real names” so they can make sure they are selling data about you and not someone else. See my Google Plus story, they told me or rather their algorithm told me my “real” name was not Google compliant. It was Chapter 52 of my series called “The Attack of the Killer Algorithms”.
“I’m Sorry Your Google Plus Name Does Not Comply With Google “Names Policies”…Barbara “Duck Algorithm” & Was Using My Real Name All Along…Killer Algorithms Chapter 52
This brings me to a great Twitter quote from Mathematician Professor Charlie Siefe from NYU who read it, saying what’s next, we will need machine compliant names? If you are pregnant and expecting you might give this some thought:) Actually since I mentioned him scroll down and watch his video about “Context” done at Google NY and you will learn a lot. He did a good radio show too called “Numbers don’t lie but people do”…
Here’s another University Professor who likes the Algo Duping coin as well, why, because it’s true.
Back on track here you also have to realize that this stuff builds on itself. So company A buys a data base and queries it with something they have in house or perhaps yet another data base they have purchased, and they sell that. Company B buys this combined “new” data base and then they take it and query it with other data bases and then they sell it. Company C gets data from both company A and B and queries it together and gee we have some new analytics for sale as they have determined by building a model that they were able to use some non linear methodologies in finding some scant relationships which may or may not really exist but what the heck, they market it and all get sucked in to the analytics having value, whether they do or do not. Now let’s not forget there’s a lot of flaws in this data, especially if one is using non credible data from social networks where folks can lie their fanny off. See how distorted and useless this data becomes when sold and resold and worked and queried. It happens all the time and this is why your credit information is so out of sorts and flawed.
So those devices, what are they selling and to who…only the shadow knows and we get one of those legal mumbo jumbo disclosures to read which basically tells us nothing and allows this to keep happening. Who knows by the time stats under a profile without your name used are done making the circuit, shoot your data could be all over the place. Let’s also not forget how data can be rematched too if the resources are available, ask any Quant on how this works. It’s done all the time and sometimes is a process in writing software with sample data bases to check accuracy of work, I used to do it when integrating software tables to make sure I was on the correct track and of course nothing was for sale there, just part of a process to ensure accurate data and queries. All developers work such processes.
If you read around the web you can see where Experian has a marketing arm that sells lists of names of expectant parents with newborns. Equifax collects salary and paystub information for about 38% of US citizens and gets that from your employers. Epsilon can sell you a list of people who donate to international charities. Drug stores are a big source of data and Walgreens made short of $800 million selling data in 2010, so I’m sure in the last couple of years that number had to rise over a billion for 2011 and 2012. Why do you think companies have so much cash, they are selling data. Keep in mind all your prescription data is not subject to HIPAA so you have open game there. United Healthcare has been making a killing for years selling this data first via their Ingenix division and now called Optum.
Now health insurance companies are buying your debit and charge card records. They stated they are using this to target those who are buying large size clothes and who knows what else. That’s all they admitted too but data in the hands of folks who can write SQL and need to keep a job as a data scientist will look for all kinds of other ways to use the information, i.e. buying candy bars, etc. You can’t really do that with any accuracy as the candy might be for your child, a friend or the homeless man in front of the drug store looking for a helping hand. Why does an insurance company need to know where you buy your gas for your car, where you eat out at and everything else that goes on debit and credit cards? This is over done by a long shot. Ever think of why your health insurance rates are so high…something has to pay for all of this and IT services and analytics don’t run cheap so you can thank all this undue data mining for contributing to higher policy charges.
Insurance Companies Are Buying Up Consumer Spending Data-Time is Here to License and Tax the Data Sellers-As Insurers Sell Tons of Data, Gets Flawed Data When Data Buyers Uses Out of Context Too
How much of this expense is really needed and how far will they encroach before nothing is sacred? Remember what is paid for policies these days and this all costs money. They are addicted in thinking that some small bit of data will develop better risk assessments. For goodness sakes we already have the MIB and others that already do a “death assessment” with estimated information on how long we will live. Yes, just what I want more device data going into those types of data bases, right? They are already incorporating some lab information. There is something to be said for predictive modeling and trending but it has to be used in context…here you have United pushing Medicare for their pie in the sky report stating billions and trillions to be saved. Nobody can predict that accurately. They get duped, refer to the videos at the Algo Duping 101 page for more on this with the videos that help explain.
United Healthcare Wants to Expand Predictive Modeling for Medicare and Medicaid–Billions in Savings Predicted in Report Only As Good as the Day It Is Published
On the other end of the marketing stick we can’t as consumers even rely on these rating sites, takes me 5 minutes or less to find errors and I barely look at them.
Flawed Data With Physician and Hospital Rating Sites- Want To Go See Michael Jackson’s Former Dermatologist? Vitals and HealthGrades Says He’s Still There - Not…Flawed Data & Algorithms Persist…
So this brings up a good question, without any laws with some real teeth and rules and a path of regulation spelled out, are all these company algorithms becoming the law? Think about it. This is the real deal folks as nothing changes until the models and algorithms are checked for accuracy and that they are not just coded to make profits.
It is also wise to pay attention to subsidiaries and acquisitions and I thought I was the lone blogger reporting on this but even the guys at Forbes wrote about all the big money United was going to make with the Humedica acquisition via their “non profit” organization they set up with using the Mayo Clinic Medical records. They have enough subsidiaries to get a cut out of almost every transaction/service today in most areas of healthcare from analytics to cheap Chinese hearing aids they distribute and sell through a subsidiary they created to do so.
Anceta, the subsidiary of the AMA is in there too as a partner, so if you read this far you know the AMA has subsidiaries in the data business too and they sell data and that’s partially how they stay alive as Humedica might be buying data from them too or vice versa. Humedica also has a banker in there too with Leerink Swann so this is posed to make money and now they have United analytics and data if they want to add more, as subdiaries of big corporations certainly share an sell tons of data.
United Healthcare Buys Humedica and Gets More Data to Analyze and Sell To Medical Device and Drug Companies–More Big Profits From Health Data
And on the other hand through all of this massive data management the AAFP found that they were, with the use of complex contracts were paying doctors in several areas of the US, less than Medicare. Actually Medicare could sue them I think if they wanted for paying less than they do, but they would have to bring in all the MDs who have signed those contracts and that would hurt the doctors.
The AAFP Confronts United Healthcare On Reimbursements, Some Are Below Medicare Rates In Parts of the US–Payment Algorithms/Formulas Calculated Deep Within IT Infrastructures Do the Job
There are good analytics when done properly and I write about those too like the one that Kaiser did with their registry on orthopedic knees and hips, now that has value to see what works best and for what types of patients. I wish we had more of this instead of the garbage we are getting now with everyone just trying to make a buck selling data. It’s time to devaluate those algorithms and ensure some accuracy and ethics with models, flat out as the data selling epidemic is out of control in more ways than one. With a PHR, I would much rather connect a device that feeds data to it, like HealthVault does with many of their devices rather than rely on a web based app for sure as that keeps me in control of what I want shared.
Consumer Mobile Health Monitoring Has Issues That Stifle Faster Growth, Epidemic Data Selling In Healthcare and Unrealistic Immediate Expectations With Patients Changing Their Behaviors
I have a friend of mine who recently filled out some health background information who is not a diabetic and listed parents as having diabetes later in life and now she’s deluged with “pre-diabetic advertisements” When does this crap stop! The states license and sell data from the public information archives and this too gets to be a problem as some states have had to put governor software in their systems to limit the amount of bots in the system and then some of the mining companies are too cheap to pay for updates at times, so do you wonder why it takes so long to get errors corrected, and of course we do this on our own dime while corporate USA profits in the billions with selling data.
Another company, sells lists of people classified as allergy sufferers and dieters and sells those lists. Be careful too as yet another company, Acxiom sells data on what you search online for a certain ailment or prescription. I think a nice DOS of everyone searching in the US for some kind of ailment would be fun, just kidding and had to say that since this is being scraped as well. How non valuable this could be when connected to data from Facebook that someone made up, right? They do it.
So those who sell products that create “saleable” data, well is there any wonder consumers are not jumping on this bandwagon? It has gone way beyond just marketing and now we have flawed data and judgments being made that are also flawed due to this entire epidemic. And it is getting worse as data scientists don’t even really know what they are doing and where value lies…watch this panel of IT executives talk about it…they don’t know what they are doing and flawed data for the sake of profit is being used against consumers in a big way.
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
We do need to license and excise tax the data sellers to get this under control as we won’t have any new jobs either until this epidemic is brought to a manageable level, plain and simple.
Time Has Come to License and Tax the Data Sellers of the Web, Companies, Banks, Social Networks..Any One Making a Profit-Latest Microsoft/Google Privacy War Helping the Cause –Consumers Deserve to Know What Is Being Sold and To Who in a Searchable Format
I know we have HHS still stumbling around and swapping spit with penalties and trying to get their analytics in better shape so we can expect little help from that end as they seem to be still searching for some black and white Algo Fairies so the FTC seems like they should get their act together big time on this issue.
Privacy Wanted–So Let’s Require Those Who Sell Web Data to Register and Tax the Transactions and Publicly Disclose Who They Sell To With a Federal Registry
We just don’t seem to have anyone smart enough in government to see what a violation of human rights this is becoming and that how much of this is in fact a waste of money but we pay for this with insurance premiums and services we receive.. Some Australian bankers made the comment that half of the analytics bought will be a total waste of investment and I agree with them on that point for sure, especially in light of what I am seeing with corporate USA behavior in thinking that every model and formula has value, they are absolutely behind the 8 ball and furthermore, why should any consumer trust any of the devices and software that sells data?
Half of Analytics Investments By Companies and Banks Will Be a Waste–What Do We Analyze with Big Data and Does It Have Value–Some Algo Fairies Would Do Better at Disneyland…
I don’t trust much of it as most of the consumer apps and software are not much more than a big rip that really doesn’t do much to help the consumer and instead pads bank accounts.
Those that don’t go to web apps to mine and sell the data are fine. I have never seen anything in the US that is so shameful and at the same time leadership that fails to recognize and act on such and continues to drive inequality as the intelligence here appears to be over the heads so so many. Keep on trying to educate patients on devices that sell data I guess, but I for one am not buying as the trade off is not worth it and would rather see jobs created where technology is used there rather than a lot of this nonsense. Nothing happens until the code that runs on the servers 24/7 is changed, so politicians can talk all they want and nothing will change, so sad state of affairs for them as well. BD
0 comments :
Post a Comment