Get best of the week

AI: Complementing tomorrow's world

The amount of data generated every day far exceeds the capabilities of human perception. And if we plan to use this volume in a more or less meaningful way, then we will have to rely on advanced technologies, including the computing power of supercomputers and the potential of artificial intelligence (AI). But while the hype around AI is focused exclusively on itself, the tools of artificial intelligence are increasingly intersecting with other technologies. This allows us not only to analyze gigantic data sets, but also to use them to solve the most serious problems of mankind.



Of course, AI has amazing potential. As more and more information flows into its algorithms, artificial intelligence is able to study and adapt them for more efficient data processing. However, we may lose sight of its true potential. Perhaps AI will be most effective where it goes beyond our own limited capabilities, making it easier for us to understand complex tasks in a way that until now was simply impossible.

Artificial intelligence has every chance of becoming the next driving factor in human progress, both in those aspects that we already understand and in what requires a deeper understanding.

Understanding the true benefits of AI


Real-world AI cases have already proven: artificial intelligence can change familiar techniques. Today, healthcare shows perhaps the greatest prospects. Scientists from the Barcelona Supercomputer Center are developing artificial intelligence models to more accurately detect retinal diseases, and their German colleagues from the Julich Supercomputer Center are modeling the work of the human brain to understand how consciousness is formed. Meanwhile, our LeHealth AI appused to detect and segment liver tumors, determine their benign or malignant. Using advanced tomography-based algorithms, it creates a three-dimensional grid for interdisciplinary analysis of clinical cases and planning operations. But why is healthcare so suitable for using AI? And what can we do to help this progress in other sectors?

Currently, the obvious advantage of artificial intelligence technology is its ability to process huge amounts of data. According to various estimates, there are currently around 40 zettabytes in the worldor 40 trillion gigabytes of data, with over 90% of them generated only in the last two years. Today, downloading such a volume of information will take about 181 million years. But, most alarmingly, only 1% of this data was analyzed. All of these zettabytes are valuable information, but at the moment we simply cannot use it with benefit, and this is where AI’s great potential lies.

However, the simple processing of such data volumes is far from the main advantage of AI technology. Artificial intelligence exists to expand and complement our current understanding of complex problems. The ability to analyze existing data and create feasible strategies is a huge help to humanity. But the real strength of AI lies in its ability to process this information, while taking into account more complex multidimensional parameters that go far beyond our capabilities in decision making.

For example, the detection of tumors. Both people and deep learning systems can analyze medical data to identify patterns, but only AI can give a meaningful interpretation. For example, our analysis of existing information should be filtered out according to a large number of parameters, including age, gender, lifestyle, occupation, as well as a huge number of other aspects that you can think of. The classification of such data and the identification of new laws by so many parameters is something that, unfortunately, we simply cannot do without the help of artificial intelligence.

We find ourselves in a kind of Polanyi paradox: understanding problems at a very complex level, we can neither algorithmize them nor solve them with the help of the opportunities that we currently have. As a result, we cannot give the machine clear instructions on how to solve such problems. However, we can give instructions to learn what we ourselves know until the machine gets a more complete picture of the problem.

With this in mind, the future of AI is not so much in developing new solutions from scratch as in improving the efficiency and analytical capabilities of existing applications.

A new approach to sensor information


The indicators that form the Internet of things, connected to a single network, are implemented in almost all areas of our lives: from sensors in the workplace that monitor employee behavior, to industrial sensors that provide constant monitoring of critical systems, as well as light, movement, temperature and control sensors waters that support the work and safety of our home. Together, these devices generate huge amounts of data, which, in turn, provide new opportunities for analysis, training and optimization. And for this, artificial intelligence is ideal.

For example, the world leader in wind energy Vestascollects data on precipitation and wind speed from 60 thousand wind turbines in 70 countries on six continents. A cluster of more than 600 nodes processes this data using AI algorithms to determine the best places to build future turbines that can reduce carbon emissions.

The role of AI in solving the most important tasks of mankind


Working with artificial intelligence can be intimidatingly complex for any organization, regardless of size. Companies often do not know where to start, they lack the experience to work with such systems. The introduction of AI is not just a way to make more effective business decisions based on analysis of market data or supply chains. The challenge also is to use AI to solve the most serious problems of society: providing energy, environmental pollution and the accumulation of waste. Combining the power of big data and artificial intelligence is key to focusing these efforts.

More articles:


All Articles