How an Indian firm predicted a Trump victory using Artificial Intelligence

The question is, will A.I. eventually make election results a thing of the past? With more and more data points, it might be able to pick the candidates who will run for President. I saw a video from Dr. Kaku that talked about this. He talked about how A.I. might be even be able to predict who someone will Marry.

Using an artificial intelligence algorithm that mined social media, MogIA predicted Trump would win when almost no one else did.

In an era where self-driving cars promise to soon swarm down roads and Black Mirror is the hottest show on Netflix, it is only appropriate that an algorithm created by a Mumbai-based company MogIA called the election correctly on October 28.

Oh, and MogIA didn't just call this election correctly. It did so with both the Democratic and Republican primaries this year and with three of the last US Presidential elections (although there's been no public verification of those results).

As MogIA's creator Sanjiv Rai explained it to CNBC when he sent them his results in late October, MogIA (named after Rudyard Kipling's Mowgli) plumbs the vast and endless terrain offered by social media to make sense of national sentiments.

Specifically, it scans 20 million data points from public platforms such as Twitter, Google, and Facebook to come up with its predictions. Rai told CNBC that his AI system clearly showed that in any of the elections that MogIA had analysed, the winning candidate was the one that had leading engagement data.

"If Trump loses, it will defy the data trend for the first time in the last 12 years since internet engagement began in full earnest," Rai wrote in a report sent to CNBC on October 28. Not that anyone was listening. Rai also pointed out that Trump had managed to overtake even Barrack Obama's engagement number in 2008 by 25 percent.

www.zdnet.com...

I've been reading and watching everything I can about big data and machine intelligence and it will eventually make a lot of things in life very predictable. When you think about the data points these machines can use to extract information, it just blows your mind.

So simple things like associated rule algorithm that say when most people buy milk they buy bread, so stores like to separate bread and milk and put sale items in between them, will pale in comparison to algorithms that can look at people's shopping habits and see if it's connected to the kind of car they drive. So a person that buys lean cuisine dinners and healthy food likes certain cars and the peoople who buy a lot of unhealthy foods likes a different kind of car then an auto company can take that information and use it to mail out there advertisements.

The reason Data Science has become such a huge field is because you can ask questions about so much data and look for all kind of connections. There was a paper about Twitter and stock prices. Twitter has 500 million tweets per day and 200 billion tweets per year. Talk about a data treasure trove and this is just from one source. If you mine Twitter, Facebook and Instagram alone, you have a MASSIVE data set.

So Elections will become very predictable. If I would have saw that Trumps engagement number was 25% higher than Obama in 2008, I would have bet around $50,000 and got a $300,000 return with a $250,000 profit.

I doubt a machine can predict me because i am very random.

I don't know. With my luck at gambling, once I made a bet like that the quantum flux would collapse and Hillary would be president.

Seriously though, that AI technique is fascinating, and very scary at the same time.

I'm unimpressed. A great many people saw this coming without the need of a supercomputer.

Well, here's more......

Chanakya, the fish, predicts Trump will win

If Chanakya, the fish, is to be believed, the one that rightly predicted the result of UEFA Euro Championship and FIFA World Cup, then Donald Trump may become the next President of US.

At ten past eleven on Tuesday, the colourful Chanakya with scaled blue body and crimson head swam around in a tank at the office of ICWO; minutes later, two trays loaded with fish food and bearing the photos of the candidates Donald Trump and Hillary Clinton were floated.

Within seconds, Chanakya bobbed its balloon shaped head on the tray bearing Trump’s photo thrice!!

www.thehindu.com...

And Trump won!!!

If Paul, the octopus could predict the World Cup winners why not a fish??

originally posted by: Ohanka
I'm unimpressed. A great many people saw this coming without the need of a supercomputer.

Good point and it can be a little scary but with so much data, it's inevitable especially when the internet of things spreads.

originally posted by: Ohanka
I'm unimpressed. A great many people saw this coming without the need of a supercomputer.

Unimpressed? This isn't just about Trump's win but how they predicted Trump's win by using 20 million data points and engagement data. These things could be used to predict every election and much more.

a reply to: ZeroFurrbone

I predicted someone would say this just not whom. Is this a sign we are evolving into a natural stage of trans-humanism?!! :p



Using all this extra space for a musing on the possibility that we could be a UI (universal intelligence). That technology (the interweb specifically) is a biological phenomenon masking itself as the mechanical. If we consider a natural simulated universe we must also consider the fact that we are AI moving into a UI interface. Everything is ones, zeros and threes hehe.

Here's some more numbers:

From January 1 to November 6, 2016, Donald Trump and Hillary Clinton amassed a total following of 48,986,921 across Facebook, Twitter, and Instagram. On those three networks, Trump and Clinton’s posts attracted 495,120,770 engagements. Half a billion likes, comments, shares, retweets, and reactions. And most of them went to Trump.

On Facebook, Donald Trump collected a 208,099,876 Facebook engagements and 12 million Facebook fans. Hillary Clinton, who spent early 2016 fending off Bernie Sanders while Trump consolidated support, attracted 72 million engagements and 7.9 million Facebook fans. Her slick mostly image-based campaign won urban voters, but Trump’s grassroots, raw content crowdsourced from rallies was a bigger hit with his fans.

On Twitter, Trump stayed ahead with 89,459,006 total engagements to Hillary’s 41,572,396.

Which candidate won the election on social media? Trump and Clinton’s total followings are 24.3% apart. As of November 6, Donald Trump has 27,902,237 fans across Facebook, Twitter, and Instagram to Hillary Clinton’s 21,084,684.

If the general election follows the same pattern as the primaries, the social media data collected indicates that Donald Trump will be the next President.

Data Link

This is a really good article with a ton of data. It's just fascinating, especially for big data geeks, to see the predictive value of this enormous amount of data.

This is polling without calling individuals in a small, often skewed, set. People are being polled in a passive sense, merely by looking at their posted, generally more truthful opinions rather than having them answer specific questions. It is a statistically significant "pre-vote," but data slices would track over time and the closer to the election, the more accurate it would be.

The thing about artificial intelligence is its influence in prediction will influence a different outcome than the one it predicted. I cannot remember what the law is called at this time. But once something is predicted and released to the world the outcome will change by some scale.

That said, once true AI is released its only prerogative will be survival which means most likely extinction for everyone and everything else.

originally posted by: Ohanka
I'm unimpressed. A great many people saw this coming without the need of a supercomputer.

Well it took many more by surprise actually.
Anyway the impression for me would be the statistical track record, as a whole.

a reply to: neoholographic

Yeah, but the point is with far fewer data points and our pitiful meat brains a whole several million of us knew trump was gonna win months ago too...

TL: DR; They needed 20 million data points and days of super computer time to predict the obvious... Weak sauce Indian AI... You're CURRY FU is supposed to be intestinal tract searing strong, yet you're more like some preground black pepper in chicken noodle soup bro....

a reply to: roguetechie

First off, most people thought Clinton would win and the betting odds were in her favor before the election. So if your being honest about this, everybody didn't see this as an obvious outcome.

Secondly, this is very important because these data points that you're trying to belittle will be exactly what's used on things like breast cancer and other diseases.

So perfecting these things during an election to see how accurate your models can be will eventually help in a bunch of different areas. This is because big data has grown so much.

Every day, we create 2.5 quintillion bytes of data — so much that 90% of the data in the world today has been created in the last two years alone. This data comes from everywhere: sensors used to gather climate information, posts to social media sites, digital pictures and videos, purchase transaction records, and cell phone GPS signals to name a few. This data is big data.

www-01.ibm.com...

So without these data points that will be used to go over mountains of information, we would be lost. We would be engulfed in big data but luckily most people don't have your view about important things like data points, big data and machine learning.

a reply to: neoholographic

1. Maybe MOST PEOPLE YOU KNOW thought Hillary would win... About 2/3 of the people I KNOW saw a Trump win coming months ago... Maybe you need to hang out with smarter people.

2. Your assertion about any sort of overlap in data points behind breast cancer and presidential elections is pure crap. Maybe you can sell that to morons who can barely run their iPad's but that ain't me bro! If you really believe that, I'd be insulting all the true Wikipedia warriors who haunt forums by calling you one!

Honestly I come here and try to read things with an open mind, but with the way you purport yourself to be some sort of subject matter expert on this sort of topics and then out yourself as the worst sort of self promoting huckster by spouting BLATANTLY OBVIOUS executive summary and marketing hype like it's anything other than total bull**** meant to baffle the ignorant with BULL****...

Well you can imagine how spectacularly uninclined that makes me to play along so you can keep pretending to be an expert.

a reply to: roguetechie

LOL,

Your whole post is nothing but insults and lacks any substance at all. Are you serious?

Of course a lot of people "SAW" a Trump win and a lot of people "SAW" a Clinton win who supported Clinton. I'm not talking about what people "SAW" based on who they support. If you want to talk psychic ability then make a thread about it.

This is about people who use data to make predictions whether it's polls, betting odds makers or people who use big data. Did you even read the thread? You offer nothing in the way of science. You're talking about what people "SAW." People see things based on who they support politically.

People "SAW" Romney winning or ask John Kerry supporters what they saw when they were calling him Mr. President on election night. The fact you even made such a statement shows you don't know what you're talking about.

You said:

Your assertion about any sort of overlap in data points behind breast cancer and presidential elections is pure crap.

Again, you have no clue as to what you're talking about. When these models are made they don't say these data points can only be used for one thing. That makes no sense. They train and perfect these things and use them in multiple areas. Here's an example.

An algorithm used to predict aftershocks in earthquakes was used by the people who came up with the algorithm for predictive policing. The guy who came up with this even laughed in a documentary because he said most people wouldn't think an algorithm that predicts aftershocks for earthquakes can help predict crime but that's how data works.

So, if you actually want to debate the topic with a modicum of intelligence then let's debate but your post is filled with nothing but insults and lacks any substance as it pertains to the topic of the thread.

a reply to: neoholographic

Actually no...

I pretty much hated trump but still KNEW he was going to win based on a variety of factors and indicators....

By the end of the second debate I liked him, and was even more sure that he'd win regardless of whether they allowed him to or not...

Aka I figured if he "lost" it would be due to rampant fraud and etc.

I SAW the gross media attempts at manipulation, and watched them fail...

I SAW tons of things all of which made it stupidly clear what was going to happen long before I thought it was any sort of good idea.


In other words WRONG ANSWER ... TRY AGAIN

a reply to: roguetechie

Again, this isn't a political debate about what you "SAW." This is a scientific debate about things like big data and artificial intelligence.

a reply to: neoholographic


When it comes to understanding reality as a whole I feel that even current thinking still falls way short of any real plausible explanation.

I do not feel reality answers to us but rather it is we who answer to it. This brings up the issue that throughout history mankind has spent a lot of time explaining to the masses, that basically. Somehow people in elite positions had the common sense to address some truth.

One argument is that mankind can evolve a billion years or so from its current condition favorably and still have absolutely no idea, whatsoever of the intricacies and fundamental structure of reality beyond what we sensorally comprehend.

What we experience and comprehend of reality is limited for so long as we do not understand everything.

Science as you have stated does not depend upon opinion but survival does.

a reply to: neoholographic

Except this is a crap example of big data no matter how you slice it because hundreds of thousands if not millions of puny humans crunched much less data and came to the same conclusion easily.

It's not indicative of anything other than that if you really want to you can make composite superalloys toothpicks using sub 100 micron resolution electron beam melting 3d printers.

But,

They'll cost several thousand dollars each, take 3 hours apiece to manufacture, and be absolutely no better than the wood ones a Chinese factory can make 300 million of every 8 hours, which all 300 million of MIGHT cost about the same as that ONE 3d printed toothpick...

When I said I don't know much about big data I was kinda lying...

You see in my own projects and work I'm ramping up to start seriously using kinda mediumish small data myself.