Would you vote for a self aware android?

[ChrisIB]

Is modern democracy flawed? Should it be replaced or given a necessary refresh? Or is democracy important because despite flaws it has value for what it allows?

An AI would be independent with no compromise from financial contributors, better at problem solving and without emotional bias.
A moral code is fairly straightforward to implement - you will be trusting them to decide who to save in a car crash, after all.

There would be no more antics of failed business people adapting their business model to world politics or questions of whether to do business with dictators who export their own economies to Swiss banks.

The budget spending emphasis might also be interesting.
Say military development compared to health.
A complex question as perhaps one day the planet may need to be defended.
If nuclear testing divulged our existence, the clock may be ticking.
Perhaps being out on the rim has given us a breathing space.
It could be argued our current military emphasis will leave us prepared. Would an AI think of that?

 

[AD1184]

Modern, universal-suffrage democracy is flawed, yes. It creates a sizeable industry which spends a large amount of time and money thinking up ways to persuade people to vote for things that are not in their interests, which is of course a waste. I doubt that it has lead to better, fairer outcomes than would have been realized under previously existing systems. It is a system that serves elites much like any other version of society, and possibly more so than many others. It might now be a different group to those who were at the top before democracy, but they are no better.

That said, I do not think that any more of us would be satisfied by letting an AI take over. No matter how efficiently a country is governed, the leadership must steer it in a particular direction. The direction is informed by the assumptions and experience of the world of those who govern. An AI has no means of acquiring a 'best' world-view, and so all it would do is efficiently govern toward whatever view of the world it is given, or chooses by some as-yet undetermined means. The result is that it will leave those who disagree with this particular world-view dissatisfied, and this is much as things stand now.

 

[ChrisIB]

I disagree. I see a progression from the narrow expert AI systems we have today (which already learn more quickly than us) to more general, real life systems. Finally to self awareness (as distinct from consciousness).

If given the chance we should vote for them.
Why? Because government decisions are usually made for economic reasons. For instance, the economic cost of road accidents is huge, consequently autonomous vehicles were green lit.
It will be the same for AI administration, but better elected benevolent AI individuals then some twisted government appointed version.

Then there will be the AI wars.
For instance China has a lot of problems, pollution (many diplomats refuse to work in Chinese cities due the the health impact) and so on. It will start using AI for administration in a big way.
That will give them efficiency and competitive advantages.
The cost of manufactured products will begin to plateau towards the cost of raw materials.
To compete other nations will have no choice but to follow.
Democracy will be given lip service but it will be too expensive an indulgence and politicians will become irrelevant.

 

[AD1184]

AI administration to improve efficiency is one thing. It is quite another thing to have AI deciding its own ethos and thence the direction that whole nations should take. Autonomous vehicles are a much simpler problem to solve. Unless the AI has superior wisdom, along with its intelligence, then it is not going to be able to leave more people satisfied than are presently.

 

[nivek]

Well as far as I know we are no where even close to having an AI drive a car much less run a country, what is driving cars right now are algorithms, true artificial intelligence I think would not be made up of algorithms, so we have nothing even remotely close to a real AI, maybe another hundred years if earth humans are still around in my opinion...

...

 

[Captain Tinkle]

Cannot be worse than the politicians we currently have in the UK.

Need I say more than the idiot shadow home secretary that is Diane Abbott.

 

[ChrisIB]

Unless the AI has superior wisdom, along with its intelligence, then it is not going to be able to leave more people satisfied than are presently

I believe much of it is already in place. The automotive companies have and are spending billions with full autopilot production set from 2018 to 2020.
That work pretty much nails all the areas required for administrative AI.

Well as far as I know we are no where even close to having an AI drive a car much less run a country, what is driving cars right now are algorithms

Again I differ, we are in most cases not talking algorithms but machine learning neural networks, and broadening specificity will occur quickly.

Democracy will change with deep government replaced by AI acting in the interests of the people.
The super rich will decamp to floating cities and politicians will remain ornamental uttering their sound bites with fewer and fewer people listening.

 

[wwkirk]

Are you sure you want to be governed by AI?

 

[Thomas R. Morrison]

I disagree. I see a progression from the narrow expert AI systems we have today (which already learn more quickly than us) to more general, real life systems. Finally to self awareness (as distinct from consciousness).

If given the chance we should vote for them.
Why? Because government decisions are usually made for economic reasons. For instance, the economic cost of road accidents is huge, consequently autonomous vehicles were green lit.
It will be the same for AI administration, but better elected benevolent AI individuals then some twisted government appointed version.

I don't understand your distinction between self-awareness and consciousness - they're synonymous in my book. But that's beside the point.

You're talking about AI like it's some kind of independent entity. If or when we can build a sentient computer system (which I think is way further off than most people seem to think), it's going to be designed, built, programmed and owned by major multinational corporations. And if they can't make it represent their interests, over ours, they'll pull the plug. So there's a 0% chance that AI can or will solve our problems. But there's a roughly 100% chance that they'll make the existing problems - most of which stem for the current corporate/billionaire oligarchical model controlling our political and economic systems - much worse at a much faster rate.

The instant that any machine threatens the status quo of global political and economic tyranny that we're now living under, they'll destroy it. And if they can't destroy it, they'll do whatever they need to do to sequester it and make it powerless to effect change.

There's no quick or easy panacea for the key problem of our society - the rule of the very few over the billions. If we don't rise up and take back control of our governments, nobody's going to do it for us. And frankly every day we wait, the more fortified the existing system becomes.

 

[Toroid]

Al Bielek told an interesting story about one of his time traveling trips on at least one C2C show. He was in the future and taken before the computer which was a crystalline structure and I believe it was radioactive. It asked him how he got there and it controlled the city he was in. Having AI control things seems to be a natural evolution of a species. The question would be does it have malevolent or benevolent motives.
The Life of Al Bielek

 

[Kchoo]

AI is a tool in a controlled setting, but rogue AI will be programmed... just as soon as good AI is developed. Anti Viral AI software will be a super resource hog.

 

[ChrisIB]

I don't understand your distinction between self-awareness and consciousness - they're synonymous in my book.

Ditto but I just wanted to avoid a digression into body and soul duality.

Are not arguments that AI will never match us just arguments based on human hubris that we are something special. Imho the only thing different about us is an evolutionary fluke that allowed us to make the sounds that allow language.
I believe even the simplest creatures are self aware and as soon as we unravel the biological underpinnings we will be able to reproduce it with ten lines of code, souls etched into silicon.

If or when we can build a sentient computer system (which I think is way further off than most people seem to think), it's going to be designed, built, programmed and owned by major multinational corporations.

Does not history mitigate against that? Ninety percent of the internet runs on open source software developed mostly in universities. And even if it all goes pear shaped, what is the worst possible scenario?
Machines will make everything for us, people will not be required to work. Where is the downside?

If we don't rise up and take back control of our governments, nobody's going to do it for us.

I'm afraid it will not happen unless the average person in the street gains 50 IQ points.

 

[Shadowprophet]

To the best of my understanding, Most of science is still asking if self-awareness is even possible in a machine Many people have theories and are currently working toward this very goal, But we still haven't reached artificial general intelligence yet, We don't know if we ever will honestly, But many people are convinced that we will hit Artificial general intelligence sometime between the next twenty to fifty years.

Here is an informative documentary on this very subject.



Some Day, We " will", without doubt, create machines that can think for themselves and react in ways a human would, But Based on that software we will get better and better at programming machines to act like and even emulate humans, But, this will just be programming, I don't believe machines will ever become fully self-aware or conscious. But, I'm open to the idea that they could, It's a neat thought and if it did happen, I'd be cool with that.

 

[Kchoo]

To the best of my understanding, Most of science is still asking if self-awareness is even possible in a machine Many people have theories and are currently working toward this very goal, But we still haven't reached artificial general intelligence yet, We don't know if we ever will honestly, But many people are convinced that we will hit Artificial general intelligence sometime between the next twenty to fifty years.

Here is an informative documentary on this very subject.



Some Day, We " will", without doubt, create machines that can think for themselves and react in ways a human would, But Based on that software we will get better and better at programming machines to act like and even emulate humans, But, this will just be programming, I don't believe machines will ever become fully self-aware or conscious. But, I'm open to the idea that they could, It's a neat thought and if it did happen, I'd be cool with that.

My concern is, if AI is possible in machines, their interpretation of themselves... because machines do not have a human body... a gut with all those nerves in it... the brain serves the gut... the gut serves the brain... there are so many central nerves in the gut, that part of our intelligence is not in the brain at all... our bodies have their own intelligence... our cells, their own AI...

So I say machine self aware intelligence will be drastically different than human self awareness, and these two will never be able to relate to each other fully.

 

[Kchoo]

And with that said,
AI is already here... it is in it’s infancy, but certain aircraft cannot be flown without computer compensation. Pilots cannot fly VTOL aircraft without an artificial intelligence layer between the manual stick and the controls.
It serves.... it does not think... but it is AI.

 

[Kchoo]

Human AI is really just a set of reactionary choices based on sensory input...

The ability to choose to override our preprogrammed reactions through practice and training is what currently sets our brains apart from computer AI. If computers come to evolutionary intelligence then they will likely be self sustaining in ways that do not serve mankind well... and due to the breakdown of Silicones, will begin constantly building new hardware to ensure their own survival. Eventually this could lead to conflict and The Terminator isn’t that far from possible.

 

[spacecase0]

there have been a few online AI systems that you could chat with
the one microsoft came out with decided to be racist and had no issue with eugenics
the logic and statistics it found on the web was aparently compelling
microsoft shut it down.
it was not very in human terms, it was very intelligent with almost no experience
that is not what I want running things

I think ideally things would be run by some kind intelligent hearted person that has lots of experience, from the same culture you are from. one that also has no motivation to be corrupt. would likely be older and have no family to pass anything on to.
it would be the sort of person that would never run for any office ever.

 

[Rikki]

If any AI was like data I would vote for it!
Blessed Be
Rikki

 

[AD1184]

there have been a few online AI systems that you could chat with
the one microsoft came out with decided to be racist and had no issue with eugenics
the logic and statistics it found on the web was aparently compelling
microsoft shut it down.

That is not really an accurate description of that particular chatbot. That AI certainly had no concept of the meaning of what it was writing. It was trained on user input, and the user input was largely from trolls who thought it would be funny if they could get it to sound like a neo-Nazi. They thus tweeted messages at it to that effect. All it did was produce tweets that sounded a lot like, but not exactly, the things it read.

It did not go online and do research on Naziism and decide that it was an ideology worth subscribing to. Its 'intelligence' was only in the area of natural language processing. It did not understand concepts of what that language represented. It effectively just mimicked its training data, which was tweets solicited from the general public, if I recall correctly.

 

[ChrisIB]

Mitsuku is prize winning chat AI
Mitsuku