Imagine if there was a super AI that wanted to improve our society, and thus started keeping tabs on politicians (from 18 years old and ahead). Every thing that was done, good or bad, is recorded and then displayed in bullet list when the time for voting draws close, that is, there will be two lists Good and Bad and underneath that will be the examples. e.g: Trump donates $x to someone in need will be recorded in Good. Assume that the AI can distinguish between minor things and major things. Also, ignore how the AI gets its information, just assume that it knows everything that the politician has done. It displays what the politicians have already done, not what they want to do or what they are going to do. It also ignores what the politicians have stated (for example, even if Trump makes a racist response, it won't be recorded).

It collects the information of current politicians, not people who want to become one. So if there is a upcoming election, the AI will have information on those politicians (not normal people who want to be politicians)

Question : Will this help? The AI wants people to not vote for politicians who are liars, which is why it has provided this list. Assume that everyone knows about this list, and that the AI can prevent others from taking it down or altering it in any way (I know it's hard to do this, but lets just assume this)

Note: The AI has just started this (it has not done this before), so there is a chance that people might take it to be a pack of lies, but the AI has provided some proof (wherever possible, like providing the names of the people involved or something) to back up its claims.

  • 1
    $\begingroup$ People are very good at avoiding/ignoring evidence against them. Also by time a AGI is created it in our world, we will be able to render indistinguishable to real voice and video of any person (with lots of photos/videos). It would like incriminating photos which are called as "photo shopped". These would be considered "video-shopped". Even with utmost order of proof, someone will resign and be given a lucrative private job. $\endgroup$ – Chinu Jul 26 '16 at 7:43
  • $\begingroup$ But who defines good and bad in this situation? Are we going by someone's personal morals? Constitutional regulations? Things contradictory to those but with governmental precedence? There are too many ways to define good and bad when it comes to these situations $\endgroup$ – Xandar The Zenon Aug 11 '16 at 4:42
  • $\begingroup$ @XandarTheZenon : Bad would be things like taking bribes, or not doing what was promised. Also, this question isn't really that good :P So there are a bunch of holes in it $\endgroup$ – King of Snakes Aug 11 '16 at 4:44
  • $\begingroup$ Just pointing out potential problems. $\endgroup$ – Xandar The Zenon Aug 11 '16 at 4:45
  • $\begingroup$ @XandarTheZenon : :) $\endgroup$ – King of Snakes Aug 11 '16 at 4:46


Contrary to Pavel, I think it would help. But possibly marginally at first.

Technical complexity

Note the technical difficulty, the AI needs to be recording all the acts of each and everyone all the time (starting from age 18). Why that? Because, if some people really show an inclination to a political carreers from an early age, most come to it a bit later. Consider this, the two candidates for the US elections were entrepreneur and lawyer before diving into politics. Ok, Clinton is a veteran of politics now, but she was 18, that wasn't sure.

In many countries there are some form of elections every year, or every two years (local, national, etc.). But let's assume that the AI aims at the national level in a country like the USA. Let's say the AI is not satisfied by the current election and possible outcome. It then decides to start recording.

Until the 2020 election, it will only manage to record 4 years of tracking. It could expand it by checking the archives of newspapers, but those are less reliable as they weren't recorded directly by the AI.

First election

Let's say the AI still did it, and in 2020 provides such a list for all the candidates to the primary elections (possibly only the Republican or the Democrats) as well as the then elected POTUS. Those only spans for 4 years.

Many people will be skeptical. Some will simply discard it: "yes, my candidate has been lying, but look what the others have been doing?" Or "aren't they all doing it?".

Then a possible next effect would be: "Yes he's a dirty liar. But the other one is simply too dangerous. I prefer a few lies than some bloody [socialist/racist/liberal/conservative]!" (pick your favourite).

Nevertheless, I think that within the primary elections, where there are more candidates with similar (enough) ideas, it might make a difference between two candidates with similar popularity and closed ideas.

Following elections

Nevertheless, the surprise that this list created, and the anouncement by the AI that the same will be repeated 4 years later... will tend to discourage candidates with too heavy past.

As the recording time and the expectation of the people increases, the effect will increase.

So more than changing directly the votes of many people, it's the politicians who will tend to present cleaner candidates.


The possible flaw, is that people and, more importantly politicians, might feel threatened by the AI. So it is very likely that they will try to deactivate it, stop it, or simply make it change its functionality.

  • $\begingroup$ I'm actually not sure about the following elections part. Especially politicians with a too heavy past tend to be self-obsessed and many people simply ignore possible negative results. Look at the three-strikes law, it hasn't reduced crime rates at all. People tend to think they don't get caught. I'd say it takes at least a mid-term run of a few decades until people somewhat realize there's no escape and even then I assume there'll still be some around thinking they can hide their past. $\endgroup$ – Otto Abnormalverbraucher Jul 26 '16 at 10:15
  • $\begingroup$ It's possible that some will stay. I'd tend to expect many of them to move to the background. You place a nice face, clean records guy... somehow controllled by a less cleaned one behind the curtain. $\endgroup$ – bilbo_pingouin Jul 26 '16 at 11:01
  • $\begingroup$ I see. I read your statement differently. Then it's actually a likely development. The most powerful people already pull the strings hidden from public sight. $\endgroup$ – Otto Abnormalverbraucher Jul 26 '16 at 13:22

No. The world wants to be deceived

Take example of USA: People feel unsafe even when statistic shows, that crime rate was going down for past 8 years.

You can use this feeling on being unsafe and inform people that this feeling comes from failed eight years of Obama's government and that it is time to change this and vote for some Republican.

The point is, we decide emotionally for our politicians

I myself was voting the Green party in last general elections in Czech Republic. Now, in some time back, the party changed their leader. And I kinda hate that guy.

There is nothing to change about Green party vision for Czech Republic and nothing to change in their agenda what they want to do if elected to Czech parliament.

The problem is, I emotionally hate the new leader of this party. So even though nothing is really changing, I am self aware of my emotional part and I want to be also emotionally attached towards my voting decision

So, long story short: Next general elections I am most probably voting for another political party, just because I cannot find enough emotional attachment for the party I voted previously.

We the people are more emotional than we admit

You can see this in heated arguments all over the internet. Even when people have enough data to decide, they tend to omit the data they do not like

Also, how does the AI knows I am joining politics?

I am 35 years old. I will continue with the Green party example.

Say I want to change things in how Green party is viewed in Czech Republic. So I join the party tomorrow.

Is the AI going to sniff through my past? Will it matter that I polluted waters when I was 19, young and stupid?

In your example, the AI has to sniff through everyone's lives once they turn 18 and keep it stored. And surprisingly everyone has to agree with that

That alone is totally unrealistic. I know you want to monitor "just" politicians, but it is totally unrealistic to assume that when someone turns 18 years, they will say: "I want to be politician. Please monitor all my steps from now on!"

  • $\begingroup$ I have edited the question, if someone new joins - then the AI starts collecting the information (lets ignore how). It ignores people who say they want to be politicians. $\endgroup$ – King of Snakes Jul 26 '16 at 7:47

What you describe is really not an AI. You're not asking it to do any thinking, just collection of data. You're also asking for it to collect data that cannot possibly be collected. It's just not possible. However, we do have a construct we use in cryptography which may be applicable: the Oracle. Oracles know things, and you do not care about how they know. Think Tyrion Lannister, only without the drinking.

In this case, your Oracle knows everything every politician ever did since they turned 18. Now what?

Every thing that was done, good or bad, is recorded and then displayed in bullet list when the time for voting draws close, that is, there will be two lists Good and Bad and underneath that will be the examples.

This is... well... daunting. Not everyone agrees on what is good or bad. Virtually everything a politician ever does in their entire career is good from one perspective and bad from another. This list has to be developed for each individual person, according to their viewpoint. This means your Oracle also needs to know your personal preferences!

Now this Oracle is sounding pretty daunting. In fact, its starting to sound more like an oracle of yore, with a direct connection to some deity. I think there may be some strong religious effects that will arise from the existence of such an Oracle.

As for the one personality trait you assign to the Oracle: "The [Oracle] wants people to not vote for politicians that are liars," I turn to Captain Jack Sparrow:

Me I'm dishonest, and a dishonest man you can always trust to be dishonest. Honestly it's the honest ones you have to watch out for, you never can predict if they're going to do something incredibly stupid.

Personally, I would take a good-natured politician that lies occasionally over an anti-christ grade evil that happens to be very good at telling the truth.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.