Technology is an interesting subject not only because it requires a lot of dedication yet a lot of people still sacrifice their time and energy to be able to understand and produce technology, but also because it always manifests itself in inter-connected and inter-related ways. The modern day notion of technology is compartmentalized in the sense that it has stages and steps of development, which produce varying different outputs at varying different times to contribute to varying aspects and applications of technology. The field of Information Technologies is a great example for such reality as there are numerous aspects of the sector that are directly connected with one another in terms of methodology and application. As information is created in the form of content online, the users form networks which are supported by servers which are controlled by enterprises and governments. As users choose to have more control over their internet or network usage, they might resort to private networks or Internet Service Providers, user-specific encryption software or a VPN service which also contribute to the general information framework. Therefore all of these technologies are also interconnected with one another and it is quite possible to find companies providing a multitude of such services in a single package in today’s IT market. The integrated nature of the IT markets are exemplary in this regard as they constitute some of the most well functioning and profitable business markets in the world which goes onto prove the assertion that modern day technology is a single unit which rests on varied and differentiated methodologies which can also be used to create different products as well.
One of the forefront players of modernity is business and therefore modern day business also is a great ground for experimentation regarding technology, especially computer related technologies. Richard Waters for The Financial Times investigates into Deep Learning within the context of modern day business to describe the methodology as “the use of artificial neural networks to carry out a form of advanced pattern recognition” where “algorithms are trained on large amounts of data, then applied to fresh data that is to be analysed.” Waters refers to a report by McKinsey Global Institute which claims that the company’s revenues would jump to 9 cents from their current 1 cent levels if they continue to implement AI based Deep Learning procedures. This leap is a result of capability to obtain and process information faster which gives a business serious competitive advantage over others. This is true for both employees with technical skills and also managers who utilize technology to understand and tackle problems. As a result, serious value is created for such business as the digital platform increases their technical capabilities significantly. As their services and products reach close-to-perfection status, customer satisfaction levels are maximized which enables such businesses to further invest into AI based Deep Learning procedures and methodologies.
One of such companies which have managed to improve its procedural efficiency and therefore its revenues is the world famous British Broadcasting Corporation, BBC. Nick Summers for Engadget reports on how the company signed a five years long research partnership with eight different UK universities. In addition, BBC will also partner with several other media and technology related institutions across Europe to create a database of observed and experienced problems which can be targeted and tackled through data analysis and Deep Learning. Such analysis and learning will be utilized to understand customer demand with respect to the corporation’s TV and online programming to construct an algorithm based system for further and future use. By developing a more personalized version of itself, the company wishes to maintain its position as the media leader of UK with a significant global presence. In addition, BBC is also investing into a new media concept called “object-based broadcasting” which makes it possible to broadcast different media productions in different formations as opposed to the current scheme of linear programming. By dissecting and re-assembling pieces of media productions, the new BBC programs will be able to create customized content while the company’s investments into machine learning will integrate further IT procedures to make the process even easier and more enjoyable for the viewer.
Developments in Artificial Intelligence and Deep Learning are making quite impressive changes in the medical field as well, making the jobs of healthcare professionals a lot easier while making the lives of patients a lot healthier. Helena Pozniak for The Guardian reports on Sam Cooper’s AI related project which the PhD student at Imperial College London is currently improving and utilizing to develop new drugs and remedies against cancer. The current state of the system allows Cooper’s team to “look at a tumour biopsy and diagnose what type it is” with algorithms giving “a more accurate diagnosis, as they are unbiased and can pick up on subtle features that are often really difficult to spot with the human eye.” With such high reliability and confidence associated with the procedure, Cooper has managed to adopt a quantitative approach to studying biological systems which has led him to understand and use AI better. As his procedure becomes more efficient over time, Cooper will continue with his investment plans to convert his ideas into an actual business as he already started a company in Toronto in this pursuit. The company focuses on automating the drug discovery process to be able to reach out to cancer patients in critical condition.
Artificial Intelligence and Deep Learning have crossed paths many times before and in the future, such collisions of interest and intent will surely repeat as well. Usman Shuja for Forbes Magazine considers this issue in his article to remind his readers that the deep neural networks utilized by Deep Learning functions are perfectly capable of solving sophisticated problems and issues, they can only do so through the provision of certain parameters. Shuja then considers the question of whether if data driven AI functions can outrun human beings in real life to emphasize that these machines and their algorithms are not capable of long-term planning which makes it impossible for them to compete with human beings regarding long-term functions they perform. The author then refers to natural language processing as a good example of this reality as several research projects have revealed that machine learning processes could manipulate language but could not replicate it and create meaning. This means that regardless of how intelligent AI systems become, they cannot develop human reasoning because they do not understand human intelligence in its purest form: language. Since the lack of language comprehension disables these technologies from interacting with human beings, their applications and roles in a supposed future human societies also become questionable, as they will most likely remain as assistants and aid as opposed to actual members of such societies.
As Artificial Intelligence develops further on, businesses and academia alike both invest into it to help aspiring and promising talents in this field find opportunities to develop their ideas and contribute to national or global economies. Jessica Murphy for BBC News reports on the recent development in the AI business in Canada, one of the most promising and emerging business markets in the Brave New World, to focus on the Vector Institute as the last innovation of such nature. The institution is based on developing techniques and methods to enhance the activities of neural networks that replicate human brain activity through computer programming. The institute’s founder, Geoff Hinton from the University of Toronto has already utilized such an approach with projects such as Texas Hold ‘Em poker where neural networks simulate human thought and strategy perfectly well. Hinton states that as data grows larger in size and computers develop more advanced capabilities, it will become a lot easier to program new applications and software to substitute and even replace human beings in real life situations and scenarios. Examples for such a reality can be found in today’s world such as speech recognition, computer vision, self-driving cars and computer players for human games such as Chess and Go. As such examples become more popular in human cultures, the business and academic interest also grow in conjunction as exemplified in the success story of Google Inc. The system is by far the most popular search engine and information sharing platform and it is almost entirely powered by Deep Learning.