Naslov
BirdbrAIn Genius

Autori
Štajduhar, Ivan

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, neobjavljeni rad, znanstveni

Skup
Riječki znanstveni mostovi

Mjesto i datum
Rijeka, Hrvatska, 08.11.2019

Vrsta sudjelovanja
Pozvano predavanje

Vrsta recenzije
Nije recenziran

Ključne riječi
artificial intelligence

Sažetak
Artificial intelligence (AI) deals with the concept of creating machines thinking and acting like human beings in a rational sense. The purpose of these machines, often called intelligent agents, is to substitute humans in performing menial work, thus leaving us humans with more time to spend on more creative tasks. By abandoning the idea of trying to mimic the human brain, as well as discarding the formal- logic approaches, feasible real-world applications of AI emerged only when the definition of AI shifted towards rational acting. Simultaneously, recent advances in machine learning, involving stacking representations and using enhanced model-optimisation techniques, gave us an end-to-end framework (i.e. deep learning) for dealing with highly-nonlinear problems, otherwise normally requiring the use of feature extraction mechanisms. This, in turn, led to an explosion of modelling representations that can be utilised for successfully automating numerous tasks, involving speech recognition from audio, language translation from text, and semantic segmentation of images, just to name a few. While automation tasks are highly important, this also triggered the development of other techniques, previously not possible, mimicking human abilities or appearance, e.g. artistic style transfer, speech synthesis, deepfake, and so on. This marked the beginning of an exciting new era involving powerful applications of machine intelligence. Moreover, AI researchers recently shifted their attention towards embedding intelligent agents with machine learning, in order to deal with the overwhelming state-space-search problems, which resulted in numerous advances in game playing. Furthermore, fully self-learning techniques were developed, breeding intelligent agents outperforming even the best players in the world. How can these findings be translated to the real world and exploited to benefit humankind, and how close are we to achieving artificial general intelligence?

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo, Filozofija, Teologija

POVEZANOST RADA