Has Anyone Heard of Swarm AI?

[KILL KILL]

I haven't until now, but this is pretty cool.

Swarm A.I. Correctly Predicts the Kentucky Derby, Accurately Picking all Four Horses of the Superfecta at 540 to 1 Odds

After Beating the Experts in Predicting the Oscars, College Football, and the Stanley Cup, a Reporter Challenged Unanimous A.I. to Predict the Kentucky Derby -- Its Software, UNU, Nailed the Superfecta in a Pick That Paid $54,000 on a $100 Bet

Unanimous spent the last two years building a swarm intelligence platform called UNU that enables groups to get together as online swarms -- combining their thoughts, opinions, and intuitions in real-time to answer questions, make predictions, reach decisions, and even play games as a unified collective intelligence. To quantify how smart these UNU swarms really are, researchers at Unanimous regularly convene swarms and ask them to make predictions on high profile events, testing whether or not many minds are truly better than one.

UNU has made headlines in recent months by predicting the Oscars better than the experts, even besting the renowned forecasters at FiveThirtyEight. UNU also surprised the sports world bypredicting the NCAA college bowl games with 70% accuracy against the spread, earning +34% return on Vegas odds. But still, the fact that average people could use UNU to amplify their collective intelligence so dramatically was met with cautious resistance.

Enter Hope Reese, a reporter from TechRepublic. Two weeks ago, she challenged Unanimous A.I. to use UNU to predict the winners of the Kentucky Derby.


"We were reluctant to take this challenge," says David Baltaxe, Chief Information Officer at Unanimous. "Nobody here knows anything about horseracing, and it's notorious for being unpredictable. Still, UNU surprises us again and again, so we recruited a swarm of volunteers through an online ad. The whole thing took 20 minutes."

Here's how it worked. During a first 10-minute session, the group used UNU to answer questions as a unified Swarm Intelligence, narrowing the field of 20 horses down to four winners. The swarm was then asked to order the four winners into Win, Place, Show, and Fourth. Then, a week later the Kentucky Derby announced the post positions of the horses, which impacts the potential outcome. So, the Swarm Intelligence was convened again, and asked if any changes should be made. One of the four picks was replaced by an alternate. This process took another 10 minutes.

The picks were then reported to the reporter at TechRepublic, who published her story the day before the race was run. The article conveyed skepticism, quoting an expert who said if this really worked, it would disrupt gambling markets.

The expert was right -- gambling may never be the same. That's because 24 hours later, the 142ndKentucky Derby was run and the four winning horses were in the exact order that UNU predicted. The odds of making such a pick, known as a "superfecta," were 540 to 1. This means that anyone who bet $100 on the published picks in TechRepublic would have made $54,000.

Cliffs:
- Guy creates AI that makes decisions based on the opinions of an online "swarm"
- AI has predicted winners of the Oscars at 70% accuracy
- AI correctly predicted a "Superfecta", at 540-1 odds at Kentucky Derby

This is pretty crazy. Is AI the future of betting?

 

[ocean size]

I am curious where the artificial part comes in when it is really groups of people making the decisions?

brb going to try this UNU thing

 

[I Am Legion]

Sky net is back.

 

[ShoelessRye]

Well in this case didn't the Derby finish 1-2-3-4 so judging by the swarm aspect, they top four favored would usually get the most support - ie the favorite getting the most bets, second favorite the second most etc - isn't that how odds work - it changes with the amount of wagers that come in?

So in reality this swarm idea would always choose favorite-2nd favorite-3rd favorite and just by dumb luck it's how this race ended.

Or I could be way off in my Derby finish order odds and disregard this post completely.

 

[Zer]

- AI has predicted winners of the Oscars at 70% accuracy

Is that racist?

 

[KILL KILL]

I am curious where the artificial part comes in when it is really groups of people making the decisions?

brb going to try this UNU thing

Because it has to analyze the results of what people are saying in order to form its own opinion. It's not simply basing its choices on a majority vote of who they think will win.

 

[KILL KILL]

Is that racist?

Probably.

 

[ocean size]

Because it has to analyze the results of what people are saying in order to form its own opinion. It's not simply basing its choices on a majority vote of who they think will win.

From the article I am not seeing that. I just did the whole UNU thing on their website, it was kinda cool but just a collective decision kinda thing.

 

[KILL KILL]

From the article I am not seeing that. I just did the whole UNU thing on their website, it was kinda cool but just a collective decision kinda thing.

I honestly don't know. If they say it's a form of AI, who am I to disagree?

 

[ocean size]

I honestly don't know. If they say it's a form of AI, who am I to disagree?

I wasnt disagreeing just curious.

Try it out on the website - the interface is like a Ouiji board, everyone pulls on it to yes/no/maybe etc. Answers, and you can see where other people are pulling to. So each person can influence the others to some degree. I think that's the artificial part, is an artifical interface that leads to decisions but ultimately it is people powered.

 

[KILL KILL]

I wasnt disagreeing just curious.

Try it out on the website - the interface is like a Ouiji board, everyone pulls on it to yes/no/maybe etc. Answers, and you can see where other people are pulling to. So each person can influence the others to some degree. I think that's the artificial part, is an artifical interface that leads to decisions but ultimately it is people powered.

Yeah, I was just on there. Would be fun to get a bunch of Sherdoggers on there on fight nights.

 

[Skip Reming]

Because it has to analyze the results of what people are saying in order to form its own opinion. It's not simply basing its choices on a majority vote of who they think will win.

It is doing what it is programmed to do. It is not thinking on its own. AI is not a good description of it.
it is still using some type of algorithm/programming.

 

[KILL KILL]

It is doing what it is programmed to do. It is not thinking on its own. AI is not a good description of it.
it is still using some type of algorithm/programming.

From what I've read, it's a subset of AI