Recently, there have been many ideas that the labor market should be “regularized” using artificial intelligence systems that will analyze and predict the demand for representatives of various specialties and provide guidelines for universities and other organizations involved in training. Thus, with the help of predictive analytics systems, it will be possible to synchronize the education system and employers. However, according to the dean of the Faculty of Business Informatics and Complex Systems Management of NRNU MEPhI, Doctor of Technical Sciences, Professor Alexander Putilov, the Russian personnel market is too young for the introduction of such systems.
According to the expert, machine learning for predictive analytics is a very powerful algorithmic method, but it requires large (and sometimes huge) databases, and actualizable ones, which is why the market is simply not ready for this yet.
“Such algorithms do not rely on “intuition” as a human person does, but they work very quickly and can analyze millions of sources of information in a matter of seconds and quickly break them down into categories”, Alexander Putilov notes. – “However, the absence of these reliable sources nullifies the entire technology, the personnel market is not yet ripe for this. It is necessary to work systematically on an array of source data, the rest is a matter of technology”.
According to the scientist, in the field of personnel development and training, we really do not yet know how to first select and then “train” employees.
“According to various estimates, more than $200 billion has been spent on the global industry of training and development, but most training specialists say that at least half of these funds were spent in vain (the developed solutions are forgotten, applied inappropriately or are simply a waste of time). However, we do not fully understand which half it is. This is a systemic problem, and it must be solved as indicated in the March decree of the President of Russia: first, the levels and terms of training, then everything else. We are now at the start of the journey, and predictive analytics, of course, will be able to help, but I repeat again – the main problem is the source data”, professor Alexander Putilov emphasizes.