Key technology agenda: What are industry leaders seeking?
Original article (in russian): https://rb.ru/opinion/kuda-smotryat-lidery/
Both Elon Musk and Mark Zuckerberg use their resources to resolve the issues related to artificial intelligence, the jeopardies and opportunities of its interaction with the human.
1. Elon Musk has set up a new company, Neura link, to create a technology, capable of connecting a brain to a PC, although a couple of years ago he said that artificial intelligence could be, in theory, more dangerous than nuclear weapons.
2. Mark Zuckerberg has been active in developing artificial intelligence-based products (in 2016, he presented a home care assistant); he also says that artificial intelligence will help our society counter global terrorism.
Yury Lobyntsev, CTO Cindicator, speaks on today’s key technology agenda.
When FinTech faced artificial intelligence.
According to Gartner’s latest report on the prospects of new technologies and the related trends, there has been an incredible interest in cognitive technologies (in a broad sense meaning partial duplication of human brain functions, responsible for the processing and analysis of information from external sources) and machine learning. Even blockchain, which seemed to become a new internet, comes after them.
It often appears that artificial intelligence only exists in futuristic movies, but it penetrated our life long ago: finance, weather, advertising, e-commerce, retail, voice control, recommendation systems, computer security, internet of things, news generators, chat bots, smart cars, education, healthcare, telecommunications, robots and art.
The finance industry, the heart of the global economy, was most happy to welcome artificial intelligence technologies.
An advantage of cognitive computer learning technologies over conservative human analysis is that machine learning is free of personal emotions and cognitive biases, which are the main cause of errors, typical for the weak human brain.
The world’s major hedge funds are using and developing high-frequency trading technologies. These projects include Sentient Technologies, Bridgewater Associates, Point72, Renaissance Technologies, and Numerai.
However, regardless of the soaring development of machine learning technologies and their application, major funds are still controlled by people with literal thinking. This is due to the fact that market life is determined by human ideas and expectations regarding this market, rather than by figures, in which artificial intelligence rationally detects a system.
Kahneman’s theory
Israeli researcher Daniel Kahneman was the world’s first psychologist who was awarded a Nobel prize in economics. In collaboration with Amos Tversky, he published his Judgment under Uncertainty: Heuristics and Biases).
In 1974, Kahneman experimentally revealed a number of non-trivial cognitive mechanisms that underlie all types of judgments and, particularly, economic decisions.
Roughly, there are two thinking systems, the rational and irrational ones. It appeared that the latter had a much higher impact on end judgment and generated cognitive aberrations in the rational part of the mind. Kahneman described these ideas in a simplified manner in his 2011 bestseller, Thinking, Fast and Slow.
George Soros’s theory
George Soros, one of the world’s most powerful financial experts, also offers his own theory. In the Alchemy of Finance, which he calls the key book of his life, he states that the market situation (the dynamics of quotes and traded values) and people’s expectations regarding the market are a single organism, rather than two separate systems.
This is the principle of Soros’s reflexivity theory. It is commonly believed that the market is always right: market prices tend to rightly compensate for future changes, even when these changes are still vague. But Soros is taking a broader view on what is happening. He believes that the expectations correlate with the course of events, but also shape future events.
People view emotions as a source of evil and mistakes in the present, although our emotions and group expectations shape our reality.
For example, any inflation expectations speed up inflation. Or take any of the latest financial bubbles, say, dot-coms or the mortgage meltdown in 2008. Regarding rational fundamental economy, there were no grounds for these events. They were triggered by group expectations.
A classical example of inflation expectations:
People, who are normally expecting a rise in the overall price level, are constantly awaiting for a new increase and buy goods for the future for fear of another price increase in the near term.
Manufacturers, who expect an increase in the prices for raw materials, equipment and components, severely overprice their goods to secure themselves. The same thing happened in Russia in 2014, when oil prices started to drop (since people are convinced that these two processes are absolutely interrelated), while currency value was growing.
Those who expected further oil price reduction queued for currency and this gave rise to further currency value increase.
Soros often repeats that the key to the mysteries of the future is in our hands, as it is in the people’s expectations regarding this future.
How can group expectations be accessed?
Only specific niche solutions and approaches of heuristic analysis can provide the uses and access to group expectations:
- marketing research;
- working with focus groups;
- surveys of individual niche experts;
- high-paid in-house professionals;
- crowdsourcing of ideas.
Still, these are minor and individual approaches, which require huge resources and a special competence of analysts in each niche. Most approaches have a high centrality factor and are therefore emotionally biased in the interpretation of judgments.
The methods of collecting expectations contain the preconditions to introduce the general imprint before the formation of expectations, which has a significant impact on the accuracy of the result by introducing the integrity of cognitive aberrations and moving away from the synthesized forecast from reality.
The existing heuristic approaches are far from being perfect in resolving the complex task of decision-making in the context of increasing indefiniteness.
To make an accurate forecast, efficient interface needs to be created to align human intelligence with the solution in the best possible way.
Suppose there is a smart interface to enable artificial intelligence access human group mind.
These could be any people, even those who are far from dealing with finance. In this case it is expectations and forecasts that matter, rather than the knowledge of the financial system structure.
The more people tell artificial intelligence about what they think about oil prices or currency value, the more accurate forecasts will be generated by artificial intelligence.
Both the number of forecasters and the assessment of previous forecasts of an individual are important to create the rating of an individual by artificial intelligence. This is very much like real life, when the Bank of Russia’s experts say there are no probabilities for the dollar rate against the rouble to go up by 20%, while in the morning we see that people started to buy currency, as a promise given by the Bank of Russia is deemed as misleading.
Such systems already exist, while we have the Internet, where human forecasts could be combined with artificial intelligence. For instance, many people of the world make financial market forecasts via mobile apps on a daily basis. People intentionally invest their attention, intelligence and time to formulate their market expectations in playful and educational manner.
Artificial intelligence collects all these forecasts in an unbiased manner, automatically follows up the accuracy and learns how to collect these forecasts on a daily basis, thus mastering the skill to synthesise the next day forecast more and more accurately.
It would be easier to rely on the opinion of competent analysts, but in 80% cases the averagely weighted reply of a crowd is more accurate than the reply of the most accurate expert. Also, this is an opportunity to find superforecasters, who can predict unexpected events, for example, black swans. Upon this, algorythmic robots receive these forecasts from artificial intelligence and complete deals on the exchange on a daily basis.
Human attention and intelligence have become a currency which can be invested and reinvested. Thus, attention could be capitalized and developed as an asset having investment attractiveness and its own ROI. Probably, human attention could enter the blockchain and become a new crypto-currency.
The world’s central think tank, the Silicon Valley Singularity University, provides lectures on the symbiosis of technology with biology, develops the theoretical concepts of limbic internet, similarly to the limbic system of the central nervous system, the brain structure, which mostly deals with the formation of complex reflexes, emotions, biological motivations and impulses.
In practice, the internet provides a symbiosis of collective human intelligence and artificial computer intelligence right now. We can observe what one may call hybrid intelligence. Who is going to be in the forefront?
These could be any people, even those who are far from dealing with finance. In this case it is expectations and forecasts that matter, rather than the knowledge of the financial system structure.
The more people tell artificial intelligence about what they think about oil prices or currency value, the more accurate forecasts will be generated by artificial intelligence.
Both the number of forecasters and the assessment of previous forecasts of an individual are important to create the rating of an individual by artificial intelligence. This is very much like real life, when the Bank of Russia’s experts say there are no probabilities for the dollar rate against the rouble to go up by 20%, while in the morning we see that people started to buy currency, as a promise given by the Bank of Russia is deemed as misleading.
Such systems already exist, while we have the Internet, where human forecasts could be combined with artificial intelligence. For instance, many people of the world make financial market forecasts via mobile apps on a daily basis. People intentionally invest their attention, intelligence and time to formulate their market expectations in playful and educational manner.
Artificial intelligence collects all these forecasts in an unbiased manner, automatically follows up the accuracy and learns how to collect these forecasts on a daily basis, thus mastering the skill to synthesise the next day forecast more and more accurately.
It would be easier to rely on the opinion of competent analysts, but in 80% cases the averagely weighted reply of a crowd is more accurate than the reply of the most accurate expert. Also, this is an opportunity to find superforecasters, who can predict unexpected events, for example, black swans. Upon this, algorythmic robots receive these forecasts from artificial intelligence and complete deals on the exchange on a daily basis.
Human attention and intelligence have become a currency which can be invested and reinvested. Thus, attention could be capitalized and developed as an asset having investment attractiveness and its own ROI. Probably, human attention could enter the blockchain and become a new crypto-currency.
The world’s central think tank, the Silicon Valley Singularity University, provides lectures on the symbiosis of technology with biology, develops the theoretical concepts of limbic internet, similarly to the limbic system of the central nervous system, the brain structure, which mostly deals with the formation of complex reflexes, emotions, biological motivations and impulses.
In practice, the internet provides a symbiosis of collective human intelligence and artificial computer intelligence right now. We can observe what one may call hybrid intelligence. Who is going to be in the forefront?
