The original inspiration of artificial intelligence (AI) was to build autonomous systems that were capable of demonstrating human-like behaviors within certain application areas. However, with the present-day data deluge (made possible by the internet), accompanied by subtle algorithmic enhancements in machine learning algorithms (leading to improved pattern recognition), modern AI systems have begun to far exceed humanly achievable performance levels across a variety of domains. Some of the most prominent examples of this reality include IBM Watson winning Jeopardy!, Google DeepMind’s AlphaGo beating the world’s leading Go player, etc. Given the above observations, it is deemed that our vision of what is to come for AI in the future need not be limited to a human imitating perspective. Instead, it may be more beneficial to build AI systems that are able to excel at that which humans have not been evolved to do or to even think about.
In this regard, one of the long-standing goals of AI has been to effectively multitask; i.e., learning to solve many tasks simultaneously. It is worth noting that while humans have not evolved to process multiple distinct situations within short timespans (i.e., in the order of a few seconds) – as interleaving more than one task usually entails a considerable switching cost during which the brain must readjust from one to the other – machines are largely free from any such computational bottlenecks. Thus, not only can machines move more fluidly between tasks, but, when related tasks are bundled together, it may also be possible to seamlessly transfer / share the learned knowledge among them. As a result, while an AI attempts to solve some complex task, several other simpler ones may be unintentionally solved. Moreover, the knowledge learned unintentionally may then be harnessed for intentional use.
With this in mind, the goal of the proposed Research Topic is to further explore the issues faced in cognitive multitasking, placing particular emphasis on computational models, algorithms, as well as new hardware advances that shall enable machines, which are free from any such issues, to be developed as consummate multitask problem-solvers. The topic shall mainly be geared towards computer scientists and computational neuroscientists, with the aim of encouraging research progress in the so far under-explored arena of enhancing the productivity of AI systems via multitasking. Key application areas of interest include the emerging internet of things, that gives rise to multiple streams of data flowing in from different sources at the same time – thereby setting the stage for AI systems that are capable of absorbing all the incoming data, processing it, and making multiple associated decisions in real-time.
The original inspiration of artificial intelligence (AI) was to build autonomous systems that were capable of demonstrating human-like behaviors within certain application areas. However, with the present-day data deluge (made possible by the internet), accompanied by subtle algorithmic enhancements in machine learning algorithms (leading to improved pattern recognition), modern AI systems have begun to far exceed humanly achievable performance levels across a variety of domains. Some of the most prominent examples of this reality include IBM Watson winning Jeopardy!, Google DeepMind’s AlphaGo beating the world’s leading Go player, etc. Given the above observations, it is deemed that our vision of what is to come for AI in the future need not be limited to a human imitating perspective. Instead, it may be more beneficial to build AI systems that are able to excel at that which humans have not been evolved to do or to even think about.
In this regard, one of the long-standing goals of AI has been to effectively multitask; i.e., learning to solve many tasks simultaneously. It is worth noting that while humans have not evolved to process multiple distinct situations within short timespans (i.e., in the order of a few seconds) – as interleaving more than one task usually entails a considerable switching cost during which the brain must readjust from one to the other – machines are largely free from any such computational bottlenecks. Thus, not only can machines move more fluidly between tasks, but, when related tasks are bundled together, it may also be possible to seamlessly transfer / share the learned knowledge among them. As a result, while an AI attempts to solve some complex task, several other simpler ones may be unintentionally solved. Moreover, the knowledge learned unintentionally may then be harnessed for intentional use.
With this in mind, the goal of the proposed Research Topic is to further explore the issues faced in cognitive multitasking, placing particular emphasis on computational models, algorithms, as well as new hardware advances that shall enable machines, which are free from any such issues, to be developed as consummate multitask problem-solvers. The topic shall mainly be geared towards computer scientists and computational neuroscientists, with the aim of encouraging research progress in the so far under-explored arena of enhancing the productivity of AI systems via multitasking. Key application areas of interest include the emerging internet of things, that gives rise to multiple streams of data flowing in from different sources at the same time – thereby setting the stage for AI systems that are capable of absorbing all the incoming data, processing it, and making multiple associated decisions in real-time.
