AI use cases in Government

We don’t want an AI President (do we?)

That is a funny notion, isn’t it?  Suppose for a moment what would happen were we to elect a truly smart, well informed Artificial Intelligence (AI) as the Leader of our country.  Immune to bribery; virtually all-seeing and all-knowing; able to instantly assess when someone is voting contrarily to what their constituents want (in support of a lobbyist with a fat wallet and an agenda).  Suddenly politicians would be answerable to the people.

Accepting bribes (or dubious “contributions”) would be almost impossible.  We’re almost at the point in social evolution where we will do away with paper money; soon everything will be in the form of electronic transfers.  It would be nearly impossible not to get caught!

It’s silly, of course, at least at this point, because we have been groomed to fear intelligent machines through movies such as The Terminator and so on.  But it is interesting to speculate what would happen if politicians were held to the highest possible standard!  What if they always did the “right” thing; if they always made choices that were for the common good?

International relations might be amazing if it knew all about a foreign country so that it could identify bargaining in bad faith.  And where AIs excel is in discovering relationships in vast pools of data.  Showing a foreign power how it could benefit by making a deal would grease the gears of international relations.

Unfortunately, it is still a bit too soon to ask for that, so…

Smaller Steps

The key to successfully reinventing politics is to employ AIs to track expenses, redundancies, wasteful or ill-conceived plans, and identifying schemes to deliberately defraud the government.  Benefiting politicians by placing job-creating industries in their home town (or at least State) would end.  Projects would be assessed logically and then located in areas where they are more efficient, with lower transportation costs for whatever they produce, and lower maintenance and production costs.

This sort of analysis would get just as much done, but at half the cost.  Such an AI could identify the most efficient way to spend money to get tasks completed, without overlap and waste.

We all know the myth of the $366 Pentagon hammer purchase, and the $3,300 coffee pot, but there is no evidence for any of the many stories like that.  Oh, granted, there was a coffee canteen (a really small food service facility installed on a military aircraft) that served coffee.  Its function was to act as a miniature restaurant, feeding the flight crew.  People repeat these stories because they want them to be true, even though there is no evidence.

The biggest culprit is called cost-averaging.  A fee is agreed upon to supply a certain number of items, say 100 each of electric ranges, refrigerators, microwave ovens, clothes washers, dryers, and dishwashers for family residences surrounding a new military base.  The item count is 700, and the contract was for $262,500.  If you divide that by 700 items then you could say that the microwave ovens were $375 each, instead of $90 at a regular store.  That sounds bad if you ignore the fact that the $600 refrigerator also was $375.  That is how these misconceptions arise.

And you can hardly blame people for re-telling these stories when we are constantly inundated with tales of governmental waste.  There is certainly no shortage of suppliers overcharging government for materials and materiel, but usually on the order of 100%, not 10,000%, so hammers and coffee pots are always exaggerated beyond belief (at least beyond the belief of most thoughtful people).

Security

A company in Wisconsin offered the chance for all its employees to be “chipped.”  That meant they could enter and leave the building with the wave of a hand; that they could pay for food in the cafeteria the same way…

This is a technology company and, when asked, 60 of the 80 employees enthusiastically lined up to have the grain-of-rice sized device injected into the space between the thumb and forefinger.  Others mulled it over and selected rings with a chip, or eventually decided to get the chip injected.  Most everyone is quite happy with the convenience.  Of course, the chips are completely removable, too.

Worried about being tracked?  The RFID chip has no GPS.  These chips are passive (no battery) and are only powered when they are in range of a transmitter and/or receiver.  We use something similar (a subset of RFID technology) called NFC or Near Field Communication when making payments with our smartphones by tapping on a terminal.  When you tap a credit card, it uses RFID technology.

Why mention this?  RFID may become a standard for identifying yourself in the next half dozen years.  Some people have already tied their Ƀitcoin accounts to an injected RFID chip so they can spend freely and never worry about loss.

Still having to enter a code when you have a chip prevents someone from trying the virtually impossible task of cloning your chip and then guessing your password.  But it has other possibilities, too.

Irrefutable Identification

Do you want to vote in an election?  You could go to a polling station if you desired, or you could log in on your home computer or mobile phone and vote from there.  Something like that, with immensely complicated encryption, would identify one person—and only that one person.

An AI wouldn’t stop there, however, since your biometrics are unique.  How you sign your name, how you type on a keyboard, how you pause when writing, what your face looks like, how quickly you blink, how your shoulders rise and fall as you breathe, how you stand with a characteristic tilt to your head…  All these elements combine to identify you, so even your twin can’t fool an AI.

Now the possibilities start to accumulate.  You can vote on any bill, legislation, proposal, or action.  Your vote would be subtracted from the total that your local representative could cast.  If enough people were interested in an idea, contrary to what their representative was going to vote, they could change the outcome.  If they agreed with that rep, they would simply let him cast the ballot they agreed with.

Keep your Witness with you at All Times

The advantages to possessing absolute identification would mean that you’d have a rock solid alibi for any situation.  You were making a purchase in a store on the other side of town at the time of the crime—charges dropped.  As long as your chip is still in your body and there is no sign of surgery, you’re free to go!  Justice, immediate and fast!

While police services are just a portion of Government supplied services, AI will certainly help them.  Programs such as PredPol (deployed in Santa Cruz) and CrimeScan (deployed in Chicago) have been ridiculously effective at helping police predict crime in specific locations around their cities, and then to increase patrols, resulting in double-digit decreases in crime rates.

Courts use it, too

Artificial Intelligence is starting to be used in some court jurisdictions to assess likeliness to reoffend, and overall risk, such as flight.  It sees the records, previous flights, or failures to attend, to help set (or deny) bail.  It helps the court set bail much more quickly.  So far N.J. and A.Z. is running it statewide, and Chicago and Pittsburg are trying it locally.

Risk to Reoffend

Some states are reporting 66% accuracy on recidivism predictions using AI programs.  That is to say that about 1/3 of released or paroled convicts have been rearrested within 1-3 years after release, just as the AI predicted.  These systems are Deep Learning systems, so they continue to refine their accuracy as long as they are running.

You can’t keep people in jail forever (in most cases), but as the system improves, so will treatment options.  Some low-risk individuals could be paroled sooner, whereas the riskier individuals could be released with an ankle cuff or other provisions in place.

Health

Another government supported service, medical care, relies on predictive analysis.  Each person is different—some respond to one treatment, while it fails for another.  Different physiology, history, or heritage could play a role in why a drug would be effective for one person and not another.  AIs can look at all available details and even guide practitioners with avenues of investigation.

AIs are superb at extracting useful data in the form or relationships that humans would likely overlook, such as incipient disease processes on x-rays or other medical scans.  Early detection of AI-enabled scan analysis is making treatment much easier, shorter, and consequently less costly.

The Takeaway

AIs look at the practical benefits of industry placement, including how it builds the overall economy and increases social strength. It will look at the impact of decisions, seeking the maximum benefit for communities, and the country at large.

We are acquiring information faster than we can analyze it—that’s just the way it is nowadays—so much of the time we are forced (if we’re altruists) to take our best guess, or (if we’re status conscious) to steer things to garner a personal advantage.

Humans can come up with very creative solutions, which AIs cannot accomplish quite yet.  We need to integrate and work together to maximize its immense analytical power; to show us where the opportunities exist to be creative; to get maximum effect for the lowest cost.  We all want cheaper government, and the best way to get that is to use the great resource to find the most efficient route to get there.

If you want to learn more, we provide webinars to show you where the future is going and how you can get there.  Don’t shortchange yourself.  Join us today!