Is Privacy necessary to Democracy?

I doubt that a dictator with AI technology will be any more evil or more deadly than earlier dictators who were quite evil and deadly without it.

Dictators rule with fear and brutality. I don't think AI changes very much.

Maybe a small number. And maybe some people will be spared because the computers don't flag them (where as an evil priest might have marked them).

I don't think this makes a big difference at all.

Greater accuracy will be a problem even if the overall number of people harmed remains the same.

It will ensure that resistance movements will be eliminated much more effectively.

I think it can make things even worse.

You didn't really answer my questions. That's a shame, because you probably have the expertise to come up with good ideas for safeguards against abuses.

I don't know what questions you don't think I answered.

I am saying pretty clearly that

1) I don't think that modern computers (including AI) present any unique danger of abuse by dictators. Everything a dictatorial regime can do with modern computers, they could do long before computers were invented.

2) The most important safeguards against dictatorship are free speech and strong civil institutions. Technology doesn't matter very much to this.

Do you have any questions that aren't answered by these two points?

Yes.

Do you think that AI technology can be abused by dictatorships?

If yes, do you think it would be a good idea to try to come up with safeguards that will prevent such abuses?

1) No. I do not think that AI technology can be abused by dictatorships in any significant way. This is because anything that dictators can do with AI they have done just fine without AI long before computers were invented. AI technology doesn't change anything significant about how dictatorships have worked for a thousand years.

2) The safeguards against the abuses of dictators have nothing to do with technology. We need to strengthen free speech and civil institutions.

I have answered your questions (although I am repeating myself because I have already given you these answers multiple times).

I disagree. I see the more-pervasive tracking of people by dictators as a problem.

This works both ways. Resistance movements (and terrorist groups) have instant communication with strong symmetric key encryption.

Interrogation by priests, or sending SS officers door to door, or having Mao's local party officials talk to family members all worked quite effectively... no computers needed.

Is there any reason to not try to develop safeguards against electronic abuses?

It depends on the specific safeguard you are discussing. Some "safeguards" cause too damage to social structure and civil liberties to be considered seriously.

Some of them are probably good ideas. We have to discuss them case by case.

This doesn't have anything to do with the fact that I don't believe AI is a serious danger to democratic society.

I'm interested in hearing your ideas.



This thread isn't about the dangers of AI to democratic society.

This thread is about the dangers of AI to people who live under ruthless tyranny.

Breaking encryption for lawful purposes is an interesting tangent. I occasionally hear in the news how Apple is refusing to break encryption on the phone of a kidnapped child or something.

While I support their willingness to leave phones encrypted for people who wish it, I'd like it if Apple would also allow people to opt out of this high security and give Apple permission to break into their phones if the government obtains a valid search warrant.

I'd gladly check a box telling Apple that they have permission to open my phone to the government with a valid search warrant. It could save my life.

In the context of this discussion, some parts of the above are true, while as a whole it is blatantly misleading:

100 years ago (your time frame), before computers, they couldn't get automated:
- activity tracking (movements, writings, interactions)
- visual feeds (activated by facial recognition when necessary)
- subject locations
- subject alerts
- communications intercepts
- and they didn't have the ability to scoop up the above date for all of their citizenry at the same time
- and they couldn't tell what people were purchasing (because it was all cash)
- and they couldn't instantly tell what bank they belonged to (because people paid cash)
- and they couldn't get automated alerts when a person said a key word on a phone
- etc

The above is obvious, and rather undebatable - the differences are quite significant. And again, you work in a field in a position to know this, so it is incredibly unlikely I am telling you anything new...so I start wondering why you keep phrasing everything in the most minimalistic way possible, rather than dealing with what is (whether or not it would ever be used for harm is another issue). It leads me to suspect that you are directly working on AI that invades peoples privacy and have to justify your involvement in such to yourself...

...which actually wouldn't bother me if you just dealt with the reality of the difference between automated computer surveillance and human surveillance.

Your current argument is like comparing the feudal lord with 1000 muskets to a feudal lord with 1000 assault rifles and trying to say the feudal lord with the muskets is going to be as militarily effective as the one with the assault rifles, or perhaps closer to your comparison...the feudal lord with horseback couriers is going to be as effective at surveillance as the feudal lord with surveillance drones. Both of those are over the top comparisons...but they do show just how much you are attempting to minimise through your comparison of 100 years ago vs now (you appear to be attempting to imply there is no difference)

Privacy of the kind you appear to want...is a thing of the past. We will never return to anything even close.

As for the Internet...if you post on the Internet, you essentially give up your privacy voluntarily.

It happens to be my opinion that the less privacy we each have...the better society is.

Vikkor... first of all, some of what you are saying is just plain wrong.

They have always been able to tell what you buy and where you go. Stores kept lists on paper, as did banks. Government agents have never had problems scooping up citizens based on religious affiliation, or banned books, or whatever. As far as listening for keywords on a phone call ... having human operators listening to phone calls is far better.

I will remind you that slavery, a brutal international trade, involved tracking and moving a cargo of millions of human slaves between countries. Human beings were subjugated, tracked, bought and sold in an institution that lasted centuries.

You are correct in one thing. Now this information is "instant", but who cares? It is not a significant change.

Whether the police gather the information they need in 2 seconds, or 2 days or 2 months... the result is the same. Government purges were quite efficient and deadly long before there were computers.

Consider the other side of the equation.

With modern computers phone calls and other communication are more secure than they have ever been in history.

With End to End Encryption any group of freedom fighters can make their communication "practically" unintelligible to any government. If I do this, the government not only can't listen in on what I say, they can't do keyword scanning, often they can't even tell if I am speaking or not.

The modern information age gives advantages to freedom fighters in fighting totalitarian regimes. Again I would argue that the end result is widely the same.

How do you figure?

If you think more information being hidden leads to a better society...go with it, IB.

I think less information being hidden leads to a better society.

Under any circumstances, it is my opinion that we have less personal privacy today than we had last year; less last year than we did the year before that; and more today than we will have a year from now.

The trajectory of that continuum holds for as far back as you want to go...and will hold for as far into the future as you are able to go.

Might as well conceive of it as something improving things...than worsening them.