AI Learns To Debate Issues....Will It One Day Run For Office?

Hopefully! It'll be far better than the current crop of corrupt thieves.

I'm sure all the voting machines will vote for it...

^{" Each side."}

^{One side is with Humans, the other side starts with DOS .}

^{Humans talk about anything. AI talks about anything, as programed.}

I'm not sure if I should say this or not. But, I guess I'll take my chances. I wonder if AI would be able to settle/prevent disputes like this:

A scientist, a clown, and a pastor walk into a cafe for breakfast. The scientist gets a plate of chemicals, the clown gets a poor excuse for food, and the pastor gets a plate of "whatever-it-is". Then, they get into a brawl. After it was over, they asked, "What in the hell was that all about?!?". The scientists said, "The pastor ate that without any proof that it was really food!". The clown said, "Both you guys liked that crap!". And the pastor said, "Oh, heaven help me....".

Totally separate research but also about debate and AI:

https://insights.globalspec.com/article/9146/researchers-develop-ai-capable-of-predicting-debate-winners

The important difference here, is that AI predicting which human wins, is different from AI actually wining the office to override human's opinions. I would hate it if human's subjective experience of life suddenly becomes irrelevant and totally objective. When do humans become no-longer-humans, but objects/statistics?

That's one of the dangers of autonomous vehicles. With a human driver, you know that the entity in control has some "skin in the game". But, with robots/machines, they don't have the same priorities.

One requirement to be president of the USA, is to be born in the USA. I think/hope that alone will disqualify AI/robots/machines from holding office. After all, nobody can even really agree what "Made in the USA" really means. Holding office is more than just winning a race/debate. It's also about the subsequent judgement-calls made afterward, that affects real people behind the "objective" statistics. Humans themselves have a problem with that. AI will only be worse, not better.

AI might be an excellent "chicken farmer". But, won't know how to deal with life-forms trying to escape the "coop/envelope/planet" or just trying to make any progress in general, at all. AI's definition of progress will undoubtedly be different than a human's definition. That's the very problem of slavery. The slave owners were more of chicken farmers, than leaders of humans.

At least it has have the ability to search information streams and quantify and qualify the information in order to determine what is true. A skill sadly lacking in the country at the moment.

Yeah, but...

...how tall is the AI?

...does the AI profess true belief in a proper mainstream JudeoChristian mythology and does it regularly attend ceremonies to that end?

...can the AI appropriately show enthusiasm without screaming wildly?

...to which party is the AI registered? Has it voted in every election since they have been legally eligible?

...how old is the AI?

...is the AI in good health, or might they be wiped out by a virus mid term?

...was the AI born here or on foreign soil? Does anyone have a copy of the birth certificate?

...what about these rumor that the AI brain was born on foreign soil?

...can the AI hold it's left hand on the Bible and raise it's right hand to be sworn in? Are we sure those hands belong to the AI? Does it count if the hands are just borrowed?

Does the AI have an acceptable spouse? Does this mean we've given up the all important battle resisting the scourge of robots marrying dogs?

.

I doubt we have an AI President in the US any time soon.

So all it needs is a good "frontman"....and be resolved to be the power behind the throne, so to speak....

Too bad - it could be an improvement.

I'm reminded of a joke where the punchline was, It must be a tragedy as it was no accident and no great loss.

Doesn't it already?

Forget 'artificial', I'm not sure there is strong evidence that even the more broad category of 'intelligence' accurately described what typically runs successfully for office in recent times.

I was more thinking along the lines of all the stuff political campaigns do to 'analyze big data', dive into the 'deep learning', and all the other trendy buzz-words of the day that 'consultants' can use to extract vast amounts of money from organizations hell-bent on getting their favored candidate elected.

Ah. 'Intelligence'. I thought you meant 'Intelligence'.

Perhaps there should be different words for those distinct meanings...

'untelligence' or 'umtelligence'?

'antelligence' or 'entelligence'?

'politelligence' or 'polelligence'?

.

_{I'm stopping now. I've written variations on the word enough to make it look weird. I found myself starting to doubt the '-gence' ending as standard.}

You never know, in your fun with the words, you might have created new terms,,

makes one really appreciate Latin... ^{a dead language}

SE thinks its not already happening...

who knew that a party was way ahead on the curve with this.

Actually, what this program does is a long ways from running for office.

It analyzes text from a report or article written by a human and finds statements that support or contradict whatever point of view it is supposed to argue. It then generates an argument. It analyzes the argument from the opponent and is able to extract points for which it constructs a rebuttal.

This is indeed impressive for an AI program. But it is accomplishing something a good student could do. And if it were fed information that was totally fictitious, it would happily create arguments that were totally removed from reality, not understanding what it is talking about.

Not ready for my vote.

It might as well be in office. Maybe it could write some laws itself.

Politicians write very few, if any, legislation themselves.

Laws are written by 1. Lobbyists; 2. Staffers

Politicians spend half of their 110 "working" days out of their offices at special cubicals in bars and restaurants close-by working the phones fund-raising. They aren't allowed to beg for money from their offices.

"... if it were fed information that was totally fictitious, it would happily create arguments that were totally removed from reality, not understanding what it is talking about...."

.

...the outcome would not be different for a human.

...the outcome would not be different for a human.

That is a good point. Humans are not immune to the GIGO syndrome.

An example that comes to mind is a book I read some time ago about "crop circles". There were some really detailed theories put forward on how UFOs could create these patterns in wheat fields.

It turned out that later, two young men confessed to creating these patterns using a board and two pieces of rope, and walking it around to compress the wheat.

https://www.smithsonianmag.com/arts-culture/crop-circles-the-art-of-the-hoax-2524283/

" In 1991, two hoaxers, Bower and Chorley, took credit for having created many circles throughout England after one of their circles was described by a circle investigator as impossible to be made by human hand.^[9]"

https://en.wikipedia.org/wiki/Crop_circle

Nevertheless, to this day, there are true believers still coming up with analyses and theories to explain.

Turns out there is already an AI virtual politician gearing up to run for office in New Zealand in 2020.

I like ganster movies and tv shows and stuff with Tom Hardy in it. I was a big fan of Boardwalk Empire and the Sopranos. I also love early 2000s garage rock (White Stripes, Jet, etc.) and british rock from this decade (Arctic Monkeys, The Vaccines, Royal Blood, etc.).

Amazon knows all of this through my purchase history, yet somehow over the course of 4 years its AI never thought to recommend the show Peaky Blinders to me. Keep in mind Peaky Blinders is a show about British gangsters doing gangster stuff with the White Stripes, Artic Monkeys, or some other rock playing in the background.

It took a friend of mine literally 10 seconds to recommend it to me when I mentioned I was looking for a new show to watch.

AI can do certain things really well, but for some reason, it has trouble with things we find easy. I think we can rest assure some jobs are safe from AI for the foreseeable future.

So what you're saying Roger is that algorithms can only go so far, Its the Human factor that can apply abstract thinking that is superior to AI and its algorithms to come up with a solution.

Which brings to mind another point. Back in the 1990's research where on to something called Fuzzy Logic for use in the algorithms to make problem solving more human like.

I wonder what happened to that... make one wonder that maybe it worked so well that programming got distracted and couldn't keep focus on long enough.

Lol, that's the problem. In order for AI to become like us, it has to become like us.