Information and Communications Technology is Merging Data Science and Advanced Artificial Intelligence Towards the Core of Knowledge Based Society -Part 3-
Romanian Distribution Committee Magazine, 2021, vol. 12, issue 3, 17-32
The paper analyses the context and the trends of the actual role/phase of information and communication technology (ICT) advances, for enabling a further progress of the Information Society (IS) towards the Knowledge Based Society (KBS), i.e., to reach a sustainable general development and in the same time avoid undesired evolutions of the Earth ecosystem challenges, like climate changes, Earth resources fading or … Covid 19 pandemic (without forgetting …food!). Observing that “evolutions” also suppose now (looking just at climate changes or Covid 19 daily consequences) a time critical race, we have emphasized the point where the ICT performances (as speed) meet the need for faster solving/processing the complex problems which actually are associated with the Data Deluge (Big Data), produced at the Earth scale, but complicatedly linking a diversity of domains. This way it is revealed the support expected from the ICT advances, where AI became the most promising instrument to face the complex problems of practically all domains, mainly by machine/deep learning (ML/DL) new models and algorithms, based now on Data Science (DS). Consequently, it is essential to deeper analyse how DS/AI/ICT could work together for better and faster results, providing the expected benefits on refined knowledge, but keeping low the undesired consequences such exponentially development could induce when facing challenges like climate changes and Earth resources fading (as most prominent and increasing every day). As concrete results of the support for AI gigantic networks, we have presented some issues regarding mechanisms and the core of the relations between the peaks of advanced processors technology and the fundamentals of their limiting factors, including the “46,225 square millimeters” chip that boasts 2.6 trillion transistors, which is in fact the maximal available (entire wafer of silicon) today. These mean that physical limits of technology and also Moore’s Law, we repeatedly have mentioned , are one of the fundamental challenges for ICT, needing revolutionary innovations in order to go on. This way it is explained why these performances are needed for the most advanced actual and future applications areas of AI, confirming the dramatic struggle of actual ICT trends, to achieve bigger chips, but, in the same time, the dimension of this permanent challenge. The benefic results of AI come along with performant processing (IT), but some of the most complicate optimization problems and eventually AI applications could be found in a diversity of areas of the communications field, although here they appear less impressive than self-driving cars industry, but it is worth to recall the crucial importance and the huge dimension of communications when converging with IT in ICT. On this line, among others, some promising AI applications, as in wireless communications and intelligent self-driving networks, are also presented. The analysis shows that the general progress of AI/ICT is more and more depending on models and algorithms, confirming also our earlier opinion that the technology advances themselves are not enough, even with the exponential development of hard components of ICT, but they have to be applied considering at every step the refined knowledge that reflects all ICT impact consequences, the fast-changing Data Deluge and the general IS/KBS context at Earth scale. In order to find the optimal matches, a lot of work has to be done on high amounts of available (but usually unstructured) data, but this could be time and energy consuming and this way we have arrived to the core of the relation between the processors performances and the targets of DL/ML/AI, where DS is coming with the expected support, including appropriate methods and algorithms. Each of these methods supposes a deeper approach, which is required to provide the progress in this complex context of AI/ICT/IS/KBS, for better results against more and more difficult actual and future problems at Earth scale. Predictive causal analytics have to consider the premises/causes of involved processes evolutions we have to predict, while the prominent actual method is still represented by prescriptive analytics, as it provides the highest results we are expecting, i.e., to have the optimal decisions after analyses and prediction of the targeted process/context. For example, this method will be benefic for the self-driving cars industry, when deciding what to do in the everchanging complicate traffic processes. All these methods are important and must work together on the general context and this is also true for both machine learning for making predictions (ML-MD) and machine learning for pattern discovery (ML-PD). The applications of ML-MD are based on previous (historical records) relevant data (supervised learning), while ML-PD are following the unsupervised models, where the ML system is looking to find hidden patterns for predictions . One conclusion is that ML-PD and prescriptive analytics could better support the most AI advances in the future more and more complex context to be optimized, in presence of the high levels of uncertainty. A deeper analysis revealed, considering the first section of the paper, that the DS/AI advances have the potential to improve the general context of ICT/IS/KBS, but the maximal results should necessarily include refined knowledge on multi-criteria optimization. This way, the link between DS/AI advances and the mechanisms of refining knowledge is naturally provided by the updated cyberinfrastructure, which could provide not only the mentioned support of the technology, but also the main level (science and engineering research) and channel to create and spread knowledge and eventually refined knowledge, with maximal efficiency. Research is identified and recognized as the top level of creating and refining knowledge at World scale, it eventually pushing its results everywhere, but the point is that, this way, DS/AI, through cyberinfrastructure, could speed up and enable more efficient the progress of all human activities. One of the most important requirements to be fulfilled in the exponential evolution of the DS/AI/ICT context, in order to get optimized solutions and refining knowledge for the actual challenges involving Big Data and other complicated problems of the Earth ecosystem is the need of standardization when developing such complex and advanced cyberinfrastructures, aiming to provide efficient interoperability and development. For knowledge refining, other features have to be also considered, because, in all cases where human mind is involved in the Earth ecosystem (knowledge must be studied in a particular context), the role of human intelligence is and must remain fundamental even in the DS/AI advances context, including the revealed link of knowledge with learning. In the same time, it is recognized that refining knowledge is a difficult and critical process of obtaining added value, where DS/AI and ICT will leverage knowledge towards the core of knowledge based society. We argued that if this knowledge is properly associated with human intelligence and innovative actions, sometimes an added value could come from it more than from technology. Using this idea and the digital transformation/disruption trends, we could witness miracle changes of the business models, based on human potential and labor but using ICT advances with unprecedented efficiency. Presenting some simple examples, another iceberg tip of ICT was revealed, as indeed they forecast a new World, which will surprise us not mainly by the technology advances, but especially by the ways ICT could leverage the innovation in the modes of impacting the human thinking and of using ICT when refining knowledge. The final conclusion is that we have to timely watch and analyse DS/AI and generally the ICT advances, in order to optimally achieve the refined knowledge that could provide the sustainable progress of IS/KBS.
Keywords: Data Deluge; Big Data; knowledge refining; data science; artificial intelligence; prescriptive analytics; machine/deep learning; wafer of silicon; Self-Driving Networks (search for similar items in EconPapers)
JEL-codes: L63 L86 M15 O31 O33 (search for similar items in EconPapers)
References: View references in EconPapers View complete reference list from CitEc
Citations: Track citations by RSS feed
Downloads: (external link)
This item may be available elsewhere in EconPapers: Search for items with the same title.
Export reference: BibTeX
RIS (EndNote, ProCite, RefMan)
Persistent link: https://EconPapers.repec.org/RePEc:rdc:journl:v:12:y:2021:i:3:p:17-32
Access Statistics for this article
More articles in Romanian Distribution Committee Magazine from Romanian Distribution Committee
Bibliographic data for series maintained by Theodor Valentin Purcarea ( this e-mail address is bad, please contact ).