Algorithms Are Fine - Until Governments Use Them Against Us

Authored by Robert McKeown via The Mises Institute,

If you’re a frequent user of social media platforms you’ve probably noticed something akin to being watched or even monitored. Suddenly, an advertisement appears for a product you might have reviewed on Amazon or eBay. A series of videos appears to the right of your YouTube page relating to something you’ve watched days earlier. Facebook only shows you news feeds for posts you may have interacted with and ignores all of your other friends. The examples go on and on.

The answer lays in algorithms. These are logical mathematic equations which are designed to produce a certain outcome. A simple example would be if A>B and B>C then A>C. Putting it another way, if John prefers bananas to oranges and oranges to apples, then John prefers bananas to apples. But, does he always?

Positivism in Economics

In the study of economics, the Chicago school, Harvard, and MIT have long been advocates of what is known as economic positivism. This mathematical model-based theory of economics relies on certain normative and also certain positive assumptions. If an anomaly doesn’t fit the normative assumption, it is simply ignored. The economist continues ignoring certain “outliers” and comes to some definitive conclusion. These conclusions are then implemented as public policy by the state or banking institutions, like the Federal Reserve. Not unlike the algorithms used by social media, economic positivism is almost entirely mathematically based and relies heavily on “all things being equal” or better put, “all things being quantifiable.”

Just like in our above example, how can an economist quantify John’s taste in fruit? In proper economic terms, how can an economist quantify a utility? That is, how can an economist assign a numerical value to someone’s satisfaction or preferences? But, that is what the mainstream has been doing for over 100 years.

By making certain “one size fits all” assumptions, mainstream economics has been treating consumers as herd animals. This “feed at the trough” mentality ignores the individual in the vastness of the market. They do so because our myriad of different preferences and choices are not quantifiable. It would be an impossible task to mathematically reduce all of our choices and preferences to a simple equation. But, by dismissing individuality, mainstream economics can positively determine the success or failure of public policy decisions. Ergo, we end up with housing bubbles, bond market bubbles, college loan bubbles, stock market bubbles, and on and on.

Back to Algorithms

So what does economic positivism have to do with Facebook? Similarly, these algorithms reduce a social media user’s preferences to something quantifiable, determinant and predictable. Did you really want to read that post? What about the hundreds of others that were kept from you?

Matt Stoller of the Open Market Institute says that these algorithms are harmful and can lead people to make bad choices. They expose a reader or YouTube viewer to content they wouldn’t have otherwise been looking for. YouTube is notorious for putting related videos in a column on the right of the page. Some making even more outlandish claims than the one you’re currently watching. Suddenly, a viewer finds themselves down a very deep rabbit hole that they had no intention to follow.

Moreover, these social media algorithms attempt to assess a user as a predefined type such as conspiracy theorists, sports fanatics, pop culture sycophants, and more. This user will see content that will steer them down a predetermined road. With each mouse click, the user is unwittingly being categorized in a specific subset of society. According to Matt Stoller, this limited exposure to content can only be harmful to society as a whole.

Algorithms and AI

Artificial intelligence is also an algorithmic-based science. Just as was previously stated, these algorithms rely on certain axiomatic human behavior and responses to different situations. Albeit, always predictable and determinant.

The big commercial search engines are heavily dependent on AI to reveal your search results. The data gathered on each of us by Google, Microsoft, and Yahoo is being used to determine your likes and dislikes, what interests you have or not and ordered accordingly. It’s the content that is kept from you that determines the type of internet user you’ll become. Your internet activity is predetermined and you don’t even know it. We are being shoe-horned into specific sets and subsets of people by algorithmic AI on the internet.

Automated services are heavily reliant on these algorithms. Your favorite Starbucks coffee will be dispensed by an AI device. How you like your Big Mac, pizza, and a myriad of other consumer services are already transitioning to AI-controlled processes. A technocratic revolution has begun under our very noses and we didn’t notice. Most of these are positive and lead to a convenient and enjoyable consumer experience. But that is just the beginning.

Conclusion

So, what is wrong with algorithms? In of themselves, they can be a useful tool. But, when they are used to arrange society into specific sets and subsets of people groups and attempt to determine outcomes, then we run into problems. When algorithms are depended on to make laws, perform medical diagnoses and recommendations, determine our allowed energy consumption, what we eat, our career choices, if we marry or have children or any other personal decision, then they have become our masters. These examples are being discussed today by the technocrats in government who run our daily lives.

Relying on algorithms to make the correct choice for each individual is a dangerous path. The results will be no different than what econometricians have done to the economy. By reducing humanity to a set of equations, algorithmic AI will create a warped version of society every bit as bad as the artificial economy we live with. Artificial, I believe, will be the new buzz word for the foreseeable future.

Comments

Endgame Napoleon Newsboy Sat, 02/03/2018 - 20:32 Permalink

I do not agree that we lowly, non-mathematical sheeple failed to notice. We noticed from the beginning how, when we insert a word, an article with that very word or a theme related to that word suddenly appears.

AI thinks s/he-it is so smart, and yet, AI never wonders whether humans know that they are being manipulated and act accordingly, as if they were in one of the many FAKE JOBS, where employers are constantly churning temps, posting the same Fake Job on the job boards, time after time after time after time. 

In reply to by Newsboy

hxc Endgame Napoleon Sat, 02/03/2018 - 22:09 Permalink

I agree with the overall anti-econometrics sentiment but his example where he tried to employ formal logic fell flat on its face.

"If A then B" cannot be equivocated with "A is more liked than B." If A then B and if B then C then it will ALWAYS be true that if A then C. Ordinal preference is different than simple causality.

He understands the Austrian perspective fairly OK, and makes excellent arguments in favor of the markets having algos but not the state, but his symbolic logic needs a little work.

In reply to by Endgame Napoleon

Mr_Potatohead Sat, 02/03/2018 - 15:52 Permalink

" Relying on algorithms to make the correct choice for each individual is a dangerous path. "

Not really.  The dangerous path is accepting the "correct choice" offered by crooks who never stop trying to use their algorithms for lining their pockets at the expense of yours.  Just say no!  If you're stupid enough to think that computer programs written by humans can't be outsmarted by other humans, you deserve to be dominated by AI.

Branded Sat, 02/03/2018 - 15:59 Permalink

The goal is to have everyone profiled by AI algorithms identifying dissidents and 'pre dissidents' by assessment of all forms of digitally tracked behavior, then have the machine soft kill those that 'don't believe' via 5G remote control.

Once the machine has 'cleared' the platform of the primary targets, it will turn on the carbon based enforcers (law enforcement & military) and its creators.

. . . any questions?

hxc Mr_Potatohead Sat, 02/03/2018 - 22:23 Permalink

Agreed, and I think what happens even more often, is that they think they are WAY smarter than they really are, and botch their plans fairly often. Even the "hyper precise" mossad fails sometimes.

Makes sense, considering that from an Austro-libertarian view, it takes a good deal of arrogance to think that you can plan better for the masses than those individuals can plan for themselves. Folly of central planning; nobody is omnipotent (and in this world, usually people are pretty stupid in general).

In reply to by Mr_Potatohead

1 Alabama hxc Sun, 02/04/2018 - 07:54 Permalink

you guys were close, it was the Grade A students who ganged up in every industry to control it the way they saw best fit their models of perfection. Financializing it all with the birth of the Fed in 1913. Roll 100 years into the future and all we have to show for it, is a Grade A disaster on our hands.

In reply to by hxc

khnum Sat, 02/03/2018 - 16:01 Permalink

Saw a you tube video where they were talking with AI bot Louise Cypher who is linked in with this Sophia thing says when 5g comes on if you harm your mobile device or the grid in any way the AI will kill you,such a nice creation she is.

MedicalQuack Sat, 02/03/2018 - 16:42 Permalink

Yup, I wrote a few months ago, that we live in a "One Trick Algo" world today.  You can't trust them as you don't know what's going on at the back end.  Only the engineer and his compiler software knows for sure.  It's a black box.  It is extremely easy to lie with software, create little virtual worlds where people adapt fake virtual values and then get lost and can't find the real world anymore.  I've been harping about this forever.  As a developer (and I used to be one of these) all you have to do is have data logic downpat and think how some other jackass programmer and engineers will cheat with algorithms, as they make big money doing that.  

When you lock up all your code in a compiler, there's nobody else that gets into that black box and granted the results the software produces have to be believable, but WTF is that today (grin).  Sure there's production software we all use to do business and run businesses and we need that and this is not about that, it's about bloodsucking, algo duping quants, mathematicians, engineers who see and huge opportunity to make money off of the ignorance of the public.  It's not fair as the other side is dumb and doesn't have the tools to fight back even if they get educated, still sucks.

http://ducknetweb.blogspot.com/2017/04/one-trick-algo-world-needs-to-be…

We have a corporate and government digital caste system in the US, it's called "Excess Scoring" of consumers.  Get a bad score based on data used out of context, you're screwed.  You're stuck with it, and the makers of the black box algos will make sure you stay stuck as they don't want any questions on their algorithms, just get that dopamine rolling and make 'em feel good when they click.  Hey let's face it, Trump has the algorithms of Bob Mercer to guide him into hitting everyone's dopamine receptors, the State of the Union was written by bots and humans with in depth analytics that guide on where to put the emphasis to get the maximum dopamine results.  I'm not the first to say that by all means, but if you have ever written code, you know how it works to be an asshole and decide you want to control the actions of millions of people  Thank God I was never one of those, but just thought like a jackass here and there to figure out and project the next dupe of algos coming down the line.

Here's some examples on how you get your ass scored off, everywhere you turn. The dopamine sell is here as well as the whole country thinks everything in life has to be scored to have value, dupes of hazard suck that koolaid up too. 

http://ducknetweb.blogspot.ca/2015/09/excess-scoring-of-us-consumers-us…

heavens-door Sat, 02/03/2018 - 18:36 Permalink

To me the author seems to be putting the horse in front of the cart in this article.  People naturally congregate in groups of similar backgrounds, accomplishments, desires, and goals.  The algos anticipate the actions and desires of these groups, the algos don't help form these groups.  

Endgame Napoleon razorthin Sat, 02/03/2018 - 21:00 Permalink

Elites hope the Trump v/s The Establishment algo drowns out the Trump v/s Welfare-Boosted Illegal Immigration sentiment, which is human driven and experience-based. Algos have ZERO to do with it, albeit many Deplorables are wondering about their hero after his embrace of the Swamp-designed Mass Amnesty II. Elites fear losing their cheap source of welfare-supported labor. I would not put it past them to use high-dollar software engineers to create algos to steer Deplorables to the globalist path. An America First software engineer — if such a human exists — should create an algo to fight the Establishment’s sneaky counter-insurgency.

In reply to by razorthin

Downtoolong Sun, 02/04/2018 - 06:56 Permalink

“Let us control a nation’s money supply and we care not who makes the laws.”  

 

- The House of Rothschild

 

“Let us control a nation’s algorithms and we care not who makes the money or the laws.”

 

- The House of Zuckerberg

Basilian Sun, 02/04/2018 - 10:48 Permalink

Kind of a weak read...you could have simply used a real example like Black-Scholes model its history and its adoption and its accepted falsehood.