UNU (Unanimous AI) is a Swarm Intelligence Platform, a form of "group intelligence" literally, "swarm intelligence" -that can be used to make accurate predictions through collaboration between several individuals.

According to the submissions of the creators of UNU, the swarm intelligence is based on the fact that the ability to predict an event by a group of individuals is greater than the sum of their individual abilities.
A few days ago, I came across a post by a Reddit user who has thought of a creative use of these tools a "EspEURImento": try to catch the 6 teams that will come out on top in their group and then the two teams that will pass as first and second in any order for each round of the European championships in 2016, and then publicly share the results that, who wants to, can use to bet.
The problem is that the UNU platform is too US-centric (the meeting schedules of the users, in fact, are designed for the evening hours of the Costa Atlantica, also hardly its US users follow the European Football).
So, instead of setting a group of UNU, CamilloBrillo-this is the name of Reditor who proposed the company-has created a Google form with 12 multiple choice questions, advising participants to behave "as if you were making your beautiful schedina the bookies. "
Voters were 343, CamilloBrillo has processed the collected data and reported the odds-that you still have time to play-with the percentages of the various winning predictions.
Ingenious: combine the great Italian passion for football with the "artificial intelligence" or with the Swarm Intelligence techniques.
I decided to contact CamilloBrillo email to ask him how it is progressing its project, finding out who owns the nickname: Andrea Nepori that deals with technology in La Stampa.
Here's what we said via Skype.
Motherboard: Hello Andrea, why you decided to groped this experiment?
Andrea Nepori: It is born of a personal interest.
Normally I am almost uninterested in the league matches, but when they start the Europeans or World become a "pallonaio ludopatico" I fixed for the games and I enjoy it to hit the various markers and coupons.
For experiment already well under way, I mentioned the preparation of this project and they asked me to write about it.
I was not even remotely crossed my mind the idea to get us out of a piece.
I never thought that a subject as nerdy could be of interest, as well, are now forced to doxxarmi alone, it had never happened before.
That said, the pretext is an article published a few weeks ago on TechRepublic, in which a reporter UNU used to predict precisely the "Superfecta" (the combination of the four winning horses in order of arrival) of the Kentucky Derby.
The combination was given by brokers to 1 to 540.
The concept of Swarm Intelligence in computing refers to models such as self-organizing systems that can perform complex action on the basis of collective intelligence.
But what stands out the UNU system than the one in computer science is properly defined Swarm Intelligence?
The concept of Swarm Intelligence in computer science is more complex than "wisdom of the crowd" - the so-called wisdom of the crowd-and refers to models such as self-organizing systems that can perform complex action on the basis of collective intelligence .
In nature, examples may be colonies of insects, flocks of birds or herds of fish.
In addition, it un'accezzione related to artificial intelligence.
UNU calculates relevant answers through a poll with a preset time limit to answer the single question.
It did not seem to notice the intervention of any particular algorithm that affinasse results.
In short, I have found that the method of calculation is tied in some special way to the platform, but rather exploits the so-called "wisdom of the crowd".
In fact, as the judgments are data anonymously, they will never be free from bias or various prejudices typical of human beings.
I'm not a pollster and do not know how to avoid bias.
So, to make the questions more impartial as possible, I tried to submit applications of Google form randomly and asked users to respond to their instinctive rifacendomi good faith-in fact, I could repeat the experience by embedding a timer in Javascript, to grant me that we do not think too much about.
And what conclusions you have drawn?
Assuming that this attempt has no scientific pretense, I wanted to see what would emerge in conditions similar to those of the platform: anonymity, speed and instinct.
A different situation from classical coupon played at the bar with friends.
Returning to our case, I had the impression that, because of the conditions for the answers, such as speed and instinct, users have indicated the teams known to be more strong and prestigious as favorites, by letting the results to a certain uniformity.
It involves taking into account past experiences: the world and Europeans have taught us that at least one of the most important teams end up doing disgusting and unlike A colder and reasoned approach, however, such as those used in so-called "systems" , one of the unexpected always end up surprised, becoming the revelation of the season.
However, there are always too many variables to predict, and the case-think-accidents can always play tricks.
The experiment ends here?
Absolutely not, I'm going to analyze the top 50 of the form responses (or 50 if chosen) to see if the amount of participants is a factor which influences the result of bias.
And then, the next step, after guess who will win the second round, will predict who will come to the quarters ...
Why did you choose Reddit as a place to bring the experiment?
For one simple reason: I am a reddittor years.
Now I attend a bit 'less, but I always really enjoy using it.
This is the ideal place for this type of evidence: a large, lively community, which responds quickly and that proved to be receptive.
The post, in fact, it went very well.
Of course, attempts to use collective intelligence on Reddit can also prove disastrous, as when it is attempted to identify the bombers Boston -in this case an innocent person had been referred to which was basically ruined my life.
But in my case-also because of the fact of treating a subject much lighter-the highest risks that could run were to receive an aggressive response, however, has not happened any of this.
Andrea, but in the end, it is putting this Italian for Europeans according to redditor?
The forecasts for Italy seem the most reliable.
Except surprises, we should be more strong in Ireland.
Belgium is a strange team, on paper looks very strong, are listed well.
In our group I worry about Sweden, I still remember when the Sweden-Denmark match, a match with the lowest prices ever, there was the so-called "Cookie" (ie, when two teams both favored tie without putting d ' agreement) and the Swedish fans had banners reading "HELLO ITALY" to tease.
Regarding the other group, however, he surprised me the confidence of redditors for Portugal, a team in international competitions has never gone well.
But the real unknown of Europeans Iceland ... in fact, it would be almost worth us a little thought.
Follow us on our new page!

From Vice