Wisdom of the crowd under computer control: how Hybrid Intelligence Cindicator is learning to predict the unpredictable

The AI race has begun: neural networks are managing to deal with and solve more and more unexpected and interesting tasks. More companies and startups are trying to take them to a whole new level by teaching them how to pass the Turing test and achieve singularity. While some lavish generous promises for those who dream of being ahead of the game (“The world’s first trillionaire will be an artificial intelligence enterpreneur”), others fear the horrors following singularity and the unpredictability of the relationships between mankind and AI. However, both stay within the same paradigm: AI can’t be avoided. The question is, will it bring us troubles, prosperity, or a complex mix of both? If yes, then in what proportions?

But there is also a third way, an alternative to a world of complete AI superiority where we put ourselves at its mercy, hoping it will be mercy indeed. The third way involves the capabilities of the human brain. Neural Interfaces (the development of which is going successfully as well, although with less aplomb than the development of neural networks) could take humans and mankind to a new level of intellectual power. At that level, we will be fairly capable of postponing the moment when we turn into pets under the watchful eye of robots and “Skynet”. We will also be capable of extending the time that we consider ourselves the pinnacle of creation, only now it will be the creation of our own making, not of evolution, gods or Engineers.

2017 has already earned its place in the future annals of human cyborgization, if only by the fact that Elon Musk, the person whose ideas always become the number one news in the world of technology, launched the company Neuralink, which plans to create “neural lace” — an implantable electronic brain lining. With e-vehicles, autopilots, Mars exploration, and new technologies in public transportation, Musk is marking the main trends of progress one after another. It doesn’t mean there was no life before Musk — just that he can give star billing to any topic he pays attention to by the mere fact of its presence within his powerful media influence.

Thus all of a sudden, Cindicator, the team of mathematicians and traders, found itself at the peak of progress. They had just been doing their job until the moment when launching Neuralink marked the incorporation of human brain features with the calculating powers and communicative capabilities of computers as one of the top trends. And Cindicator, suddenly even to itself, happened to be among the leaders of that trend.

What does Cindicator do?

Cindicator is a platform which combines the capabilities of the collective intelligence in the most direct sense (the so-called “wisdom of the crowd”) with the analytical power of computer systems in order to predict future events, first and foremost, in the financial sector.

Wisdom of the crowd is an effect that has been noticed and proven multiple times. It occurs when the average number of results of guessing by any large number of people turns out to be close to the correct answer. Remember Raymond Babbitt (played by Dustin Hoffman in the movie «Rain man») who managed to count all the toothpicks spilled by the waitress at a glance? A common person is incapable of doing so, however, if a hundred people put their guesses down after giving it a glance, then the average number would be close to the truth.

If we start from afar, any mass expression of will from the crowd, when carried out correctly, is indeed the expression of the collective unconsciousness, which can be eternalized: the voice of the people is the voice of God.

But stepping out beyond amusing lab experiments and finding a practical application of the “wisdom of the crowd” hasn’t been managed by anyone yet. First, because of understandable logistical reasons: the technical capability of involving unlimited large numbers of people in solving tasks this way has only recently appeared. The smartphone and mobile internet boom made the “crowd” fully available for those who are interested in its “wisdom” — and let the guessing process take the form of a game.

Secondly, according to the mathematical principle of “a problem well stated is a problem half solved”, one should have a knack for asking questions of the collective intelligence, so that the situation doesn’t turn out like in The Hitchhiker’s Guide to the Galaxy:

The answer to the main question of life, the universe and so on: I thought it over quite thoroughly. It’s 42.

Those who study wisdom of the crowd say, that it is best used in mathematical or geographical tasks — tasks with definite clear-cut answers. In other words, places or numbers. Since a place is just a system of coordinates, it all comes down to numbers. The paradox here is that few of us were math fans at school, however, collectively we turn out to be decent accountants. It’s interesting — what will happen if students start taking exams by way of voting all together for these or those answers?

Thanks to the “connection” to the collective intelligence via the mobile app, Cindicator reaches full decentralization of its collective mind. This is one of the main conditions — participants must avoid any communication and exchange of opinions within the group, in order to avoid the influence of some individuals on others.

How to earn on forecasts

When the character from “Back to the Future” decided to earn some money by knowing the future, he decided to bet on sports. Cindicator, choosing an area to apply its efforts, started with the financial market, an area that has given the most promising feedback. It connected an urgent daily need in forecasts (what is there to say when individual financial tools by the same name claim to require Nostradamus), and accurate feedback, allowing them to evaluate the quality of the forecasts, and, of course, giving the opportunity to effectively monetize incoming data.

Strengthening the team of mathematicians and data-analysts with traders with serious experience in managing high positions, trading in global stock, futures, foreign exchange, and cryptomarkets, Cindicator has started finding its own “special path” in the brutal world of trading.

In reality, practical application of the wisdom of the crowd in trading at stock and foreign-exchange markets is a bit more complicated than daily attempts to collectively guess tomorrow’s dollar exchange rate. Cindicator constantly tests various types of questions, data, new trading strategies, and models for optimizing the usage of collected data for creating trading robots and analytical products (indexes, price levels, market sentiments etc.). It constantly carries out backtests and forward tests to adapt those models to conditions of constantly changing markets.

Besides working with the results of the crowd’s forecasts, Cindicator also aligns the forecasting machine itself. As the experience of Cindicator shows, fine alignment could likely speed up its engine effectiveness by multiple times. One example of this is a joint experiment with the Moscow Exchange, during which a Cindicator trading robot made trades for 3 weeks based on the collective opinion of 863 app users, 40% of whom had never traded at a stock market before. Overall, the robot modeled 27 trades, 17 of which were profitable. As a result, within the time limit of the contest, the model portfolio increased by 2.8% in USD (47% per annum).

To increase its performance, Cindicator models need to analyze not only the accuracy of each collective forecast, but the effectiveness of each of thousands or tens of thousands of forecasters, its strengths and weaknesses — that big data, which has been widely discussed in recent years, but has hardly been touched.

Cindicator’s job with big data includes the system and methods for deciding the forecast significance of each app user based on a personal time-table for various types of questions (time spent making the decision, forecast history for each type of investment tool, connections between answers, type of analysis in use, etc.), dynamic feedback after each trade which considers both the fact of profit or loss, as well as its size at the end of each forecast.

The mind’s sleep gives birth to monsters. Analysis of the collective intelligence — superforecasters

Thus, the computing power of Cindicator does more than just align the machine of collective forecasts. It singles out individuals from the crowd. At first glance, it might seem at odds with the idea of the wisdom of the crowd, when large amount of noise, mistakes, and subjectivism is compensated by heterogeneity and multidirectionality (a simple example is the Gaussian distribution). However, the experience of Cindicator shows that slight dirigisme, carried out with high mathematical accuracy, allows not only the alignment of the spontaneous power of collective forecasts, but also (something that couldn’t be expected earlier) the selection of superforecasters — individuals who turn out to be stable forecasters by themselves.

Superforecasters — superheroes of the real world

Until recently, the “wisdom of the crowd” phenomenon remained only slightly studied. A decent number of experiments have been conducted to prove its existence. But there have been no calculating or organizational powers to dig deeper.

Cindicator is not a research project. However, a side effect of the practical application of the “wisdom of the crowd” is a deeper understanding of the principles of this phenomenon — and new discoveries. New discoveries, like the proof of the theory about those superforecasters (superpredictionists — people who, in most cases, give more accurate forecasts than others) right after the launching of the platform. The number of superforecasters among all Cindicator users is about 2%. Further analysis showed that even they are not homogenous — some superforecasters are universal as predictionists, while others have on average more accuracy at predicting certain types of events. Reasons for the uniqueness of this 2%, as well as the specialization of the others, still have to be determined. It could be a natural talent, or a competence coming from experience. In the aforementioned example with the Moscow Exchange, the author of the most accurate forecasts was an investor with 10 years of experience. Yet, it doesn’t devaluate the idea: his personal trading result could be +1,48% interest within 3 weeks (25% interest per annum), which is twice as low as the profit given by the collective wisdom of the Cindicator forecasters (2,8% in USD within 3 weeks/47% interest a year).

In comedy shows there is often a gag, when a housewife guesses the results of sport matches better than her husband, who is a fan of this sport. In real life, the least expected people could have a forecasting talent in predicting currency exchange rates or future weather conditions, not even having a clue about it, neither having a clue of what they could accomplish by using this superability.

Ahead, to noosphere

Cindicator calls the wisdom of the crowd corrected by machines “Hybrid Intelligence” (HI). One can’t call the project team ‘developers’ since they develop only part of the system. Its main force is, literally, alive. So the Cindicator could be more accurately described as ‘administrators of hybrid intelligence’, those who stand at the beginning of its exploration, studying and taming.

If we combine Cindicator’s knowledge about the wisdom of the crowd and experience in directing spontaneous power of the collective intelligence towards the solution of applied tasks together with developments of, say, Neuralink, if Elon Musk is followed by the same success in those achievements as with e-vehicles, then hybrid intelligence Cindicator, in combination with a neural interface, could ensure a breakthrough for people to new technological, intellectual, and social levels, which was earlier exclusively related to the invention of the AI.

The physical (or better, bioelectronic) interface created by Neuralink will require software developments, which will allow us to carry out emerging opportunities. Likewise, the modern smartphone without a rich set of programs is nothing more than just an expensive flashlight. Hardware development is more complicated and expensive, but the one who has the best software wins in the end.

Applying its technologies on Neuralink’s platform, or on any other similar project, whichever comes to a finish first, Cindicator will be able to transcend the barrier of conditionality in determining “hybrid intelligence” and create a hybrid of a human brain with the calculating powers of machines without any quotation marks — a true hybrid intelligence. In fact, it will mark the transfer to noosphere. And so far, out of all available alternatives, the abilities of the human mind, developed by integration with electronic brains and united, perhaps, into something like a super-mind, seem like the best protection from AI threats, both real and imaginary. At least it will keep the possibility of the human mind to cognize reality alive, thus postponing singularity for some time — i.e the moment, when the speed and complexity of technological development goes ahead of the human’s ability to realize all that’s happening. Hybrid Intelligence is, so far, our best chance to maintain our top position in the food chain.

Hybrid Intelligence for everyone: why Cindicator is going to Token Sale

Actually, let’s get back to the present. Right now, Cindicator is just a “hybrid intelligence” and those who control it are doing things any one of us could do if one had a working AI — using it for financial markets and trading. Chosen at the very beginning, the standard b2b business model of Cindicator turned out to be of low scale: during pilot cooperation with huge hedge-funds and banks, it became clear that their interest in using Cindicator was accompanied by a mandatory condition of monopolistic access to the full capabilities of the platform or at least exclusive access to individual data.

To avoid future conflicts between clients, the Cindicator team made a decision to regulate at the next level of its development the number of funds receiving full access to the capabilities of the hybrid intelligence. The users of HI will get an opportunity to maximize the effectiveness of its usage in selected areas — at cryptomarkets, traditional stock markets, derivatives, venture market, etc.

The new period in development of the Cindicator company starts on September 12th, when the launching of the open Token Sale of CND infrastructure tokens will take place. Successful distribution of these tokens will solve the following important current and future tasks:

1. platform and infrastructure development of HI;
2. development and release of new products;
3. stimulation of the exponential growth of the number of forecasters on the platform;
4. replenishment of the existing forecaster motivation fund;
5. continuation of the scientific work;
6. payment for marketing and legal expenses.

After conducting the Token Sale, Cindicator will create an additional dynamic fund for motivating forecasters, which will be attached to the profitability of the experimental trading portfolio. This additional reward will be distributed proportionally to the accuracy of the analysts’ forecasts and the result of transactions, conducted with the help of data, received from the collective intelligence of platform forecasters. Results of this trading will be added to the HI transactions portfolio in order to boost the interest and demand for Cindicator products among professional financiers and traders: trading indexes and indicators, price rates and other types of predictive analytics of traditional financial and cryptomarkets.

The release of CND infrastructure tokens is the next step in creating a technological infrastructure (API + forecasting module + data science module + selling module + GUI module) that will be used by investment funds of a new format and utilizing all the products and powers of hybrid intelligence technology as efficiently as possible.

Funds that purchase this technology (apart from purchasing tokens at market price from their owners at stock markets) will be regularly giving a percentage (performance fee) from its potential profit, with which token buyback from the market will be carried out in order to increase motivation among all active participants of this ecosystem (forecasters, traders, data-scientists). Availability of that infrastructure for the funds is planned for 2019.

The more interesting the project itself is (in comparison with the whole mass of ideas, which has swarmed the information space with their monotonous and one-time coin offerings), the more interesting the list of products, which will be given to regular CND token owners, becomes. Attached to CND tokens (depending on their quantity), are bonuses, such as access to indicators, ratings and internal analytical products of Cindicator, including:

– indicators of traditional and cryptomarkets (probability of increase/decrease in asset price, possibility of anticipating analysts’ consensus at corporate and macroeconomic events, indicators of reaching certain price levels, indicators of the likelihood of significant events influencing the market);

– additional service products for trading (Telegram and Slack bots, notifications, products for monitoring the portfolio condition);

– analytical products (ICO ratings, analysis of the market condition, due diligence ICO, analysis of investor’s portfolio);

– market indexes and sentiments, generated by HI.

Thus, CND token holders will be able to use the intelligence capabilities of Cindicator for more effective management of their other crypto- and regular investments.

Release and sale of Cindicator tokens starts on September 12th. You can try your forecasting skills right now by installing the Cindicator app for Android and iOS. Both current and future users, who become active forecasters before the beginning of ICO, will get an option of special bounty-reward upon its completion.