4 minute read

Intelligence Augmentation

The evoluTion and The fuTure of arTificial inTelligence

By constantinos valanides, Senior advisor ii, KPMg limited.

NNowadays, artificial intelligence (AI) has become an integrated part of our everyday life. Personal assistants on mobile phones or autonomous vehicles may prove that AI has managed to reduce human intervention and improve the driving experience respectively and such solutions are undoubtedly simplifying our lives on both a personal and professional level. Human intelligence can potentially be enhanced and supported by AI in the work environment. A subset of AI, Machine Learning (ML), has the capability to consume and analyse huge volumes of data and provide users with useful insights. ML algorithms uncover hidden patterns in the datasets and provide accurate predictions, allowing us to take important and informed decisions quickly. Such algorithms can be a part of automated procedures or embedded within technological solutions. In the near future, autonomous systems will have the ability to learn from the cognitive behaviour of humans and demonstrate approximately the same behaviour under different circumstances and parameters. AI techniques can be better developed in a structured environment, where the necessary information and objectives are well-defined. In such cases, the accuracy of an AI model is significantly higher than before and it can exceed human judgment in specific cases (narrow AI). The classification of e-mails as safe or malicious is an example of this type of AI. On the other hand, if the required information or the business objective is not completely clear, Intelligence Augmentation (IA) comes in to play a key role. Human intelligence can be enhanced by technology through the IA approach. The human has a central role in the interaction with the machine. For example, a car collision avoidance system warns the driver to avoid an upcoming car accident with the provision of advanced information to enable proactive action. However, in this case, the driver still acts as the main user of the system by utilising the insights from IA to prevent the accident.

The difference between IA and AI is that IA aims to enhance a human’s level of intelligence so as to take an action or decision, being wiser than before, using similar technologies to AI. On the other hand, AI tries to learn and mimic human cognitive behaviour. As an example, the manager of an organisation’s marketing department can be made aware when there is high probability of a specific group of customers leaving the organisation. At the same time, IA technologies support users by recommending a series of potential actions to prevent this negative event. Furthermore, IA can successfully combine the results of AI algorithms and recommend specific products to a group of customers, considering both the proper timing of such a move and their personal needs. The customer experience is thus enhanced and the customer lifecycle is extended. A large number of financial institutions are already adopting AI technologies to support their operations. In some cases, they use advanced algorithms to identify relationships and trends in the data. This enables them to move from the descriptive analysis of data to a more predictive and proactive approach. For example, algorithms like these can detect fraudulent transactions based on a large number of parameters and reject them even before their execution. Another use of AI within financial institutions is the estimation of the credit score of their customers to a high degree of accuracy, which can lead to the minimisation of the institution’s operating risk. Furthermore, AI is introducing new channels of communication with customers, such as Chatbots. Chatbots incorporate various Natural Language Processing and Generation algorithms in order to be able to understand and communicate with customers in a similar way to humans, thus saving time and costs. Organisations that adopt a series of AI algorithms may have a number of different results most of the time. It is up to the user to understand and utilise the results and take a decision or an action. IA can easily recommend actions derived from the results of AI. It can also estimate the outcome of an action, giving the user the opportunity to select the one with the most profitable result. These examples demonstrate the important role of AI and IA in financial institutions for both employees and customers. IA has a crucial role to play in the enhancement of human intelligence. Users are provided with combined information derived from different AI models and data sources that were not available before. Based on these new insights enabled by IA, humans have the opportunity to take better informed and, as a result efficient, more profitable decisions. There is no competition between AI and IA but both will have a key role in automation in the near future.

IA cAn eAsIly recommend ActIons derIved from the results of AI

0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1

This article is from: