The potential of AI

Although the term artificial intelligence was coined as early as 1956, AI has grown in popularity these days due to the collection of more data, the creation of advanced algorithms, and improvements in computing power[1]. Previous technological revolutions were mainly based on automating manual labor activities. Today, we are talking about generative artificial intelligence, which will likely have a breakthrough impact on knowledge-based work and the need to make key decisions based on it.

Artificial neural networks

Generative artificial intelligence models mimic the actions of biological neural circuits. An artificial neural network (ANN) is a number of interconnected neurons. The output of the first neuron passes a signal to the input of the next one, this one processes the signal and sends it to the next one – and so on... An artificial neural network is a powerful tool for artificial intelligence learning. Speaking of learning, a certain term in this field is necessary, which is deep learning, or deep learning.

Deep learning

One of the techniques of machine learning (ML) is deep learning. Some deep networks are able to learn without the need for labels or human supervision. Artificial neurons are arranged in several layers. The outer input layer receives stimuli and passes them to subsequent layers. As in humans, after the information passes through several levels, it goes to the "brain" of the machine. In this way, from a simple stimulus received by our senses, the information "hurts," "warmth" or "familiar face" appears in our brain. In turn, a number is fed into the "brain" of the machine. Depending on its value, the machine acts accordingly, such as recognizing an object or performing a specific action[2]. Current deep learning models, unlike previous ones, can process extremely large and diverse sets of unstructured data and perform more than one task.

The potential of generative artificial intelligence

Traditional advanced data analytics and machine learning algorithms are very effective at observing and classifying patterns. Generative AI, however, pushes the boundaries of creativity and problem– solving. The latest tools based on generative AI can perform a range of routine tasks but also write texts, compose music or create graphics.
Generative AI has a remarkable ability to interpret open-ended commands, write, summarize, code, brainstorm, and remix all the ideas and skills that internet users have demonstrated over the past 20 years.
The technology has found Immediate application in areas where people spend a significant amount of time reading and writing, with the goal of improving information gathering and synthesis. Using generative artificial intelligence, organizations are seeking to optimize productivity and revolutionize the way they process and collect information.

Elements of generative AI systems

A generative artificial intelligence system consists of three main components: a generator, a discriminator, and hyperparameters. The generator creates data samples, and the discriminator evaluates them based on previously observed data. Hyperparameters are variable parameters that define the performance of the generator and discriminator, determining their weighting – that is, the degree to which a particular parameter is or is not relevant. These three components work together to train the generative artificial intelligence model to produce high-quality data that resembles training data, i.e. carefully selected and cleaned information[3].

Value for business

The business potential of generative artificial intelligence lies in its ability to automate and streamline creative processes, thus saving companies working time and resources. Therefore, we are facing a significant change in the workforce, the production processes in the economy, the tasks performed and the skills needed to succeed in business are also being transformed. Given the ability of generative AI to deliver results in a wide range of formats (text, image, code, and others), the possibilities are endless. The key, therefore, is to understand the ways to use it that would bring the most value to a given industry.
By learning from large data sets, generative AI helps create new content or reduce the time and cost of conceptualizing it. In addition, it helps make data more useful and easier to interpret.

Reducing the workload of the employed

Artificial intelligence, in cooperation with other available technologies, makes it possible to automate the activities of working people, which currently consume 60 to 70% of working time[4]. Until recently, this figure was close to 50%, but the streamlining of duties is largely due to the increased ability of generative artificial intelligence to understand natural language.
Generative artificial intelligence, while collaborating with employees and streamlining their work, simultaneously increases their productivity[4]. It reduces the burden of routine duties while allowing them to focus on more demanding, creative tasks.

The permanent link between finance and technology

The financial market nowadays is inextricably linked to technology and has therefore become dependent on it as well. One can venture to say that this area is no longer able to function without a technological background. Virtually all key processes of the financial sector have been transferred to the digital domain. Mostly, however, AI remains behind the scenes, optimizing business functions or making recommendations on the next product to buy.

The pillar of the financial market is the customer, both individual and institutional, who is provided with access to products and services to manage their assets.

Innovative customer service

The chatbots created within the company are able to provide immediate and personalized answers to complex customer inquiries. So once again, the barrier of the native language or location of the person in need is transcended – voice assistants and chatbots are available anywhere and anytime, speaking a wide range of languages. They improve the quality and efficiency of interactions, while affecting a positive consumer experience. As a result, uncomplicated problems are solved with the help of automated tools, and customer service teams will focus on interactions that an algorithm cannot handle.

Thus, technology directly influences customer satisfaction and satisfaction, which results from a positive experience with the company. Thus, it fosters long-term relationships with consumers, enabling a better understanding of their needs and providing more personalized recommendations in the future.

Research and product development

The use of generative artificial intelligence models both reduces research time and improves subsequent simulation and testing. It is noted that this technology is capable of yielding productivity gains worth 10 to 15% of total research and development costs. Generative planning can also enable improvements in the designs themselves. What's more, it also contributes to lower production costs by helping to select more efficient materials. As a result, it makes it possible to optimize designs, reducing costs associated with manufacturing and logistics processes. As a result, the quality of products offered by companies increases.

AI in IT, or happy developers

A McKinsey study indicated that equipping programmers with the tools they need to achieve maximum productivity significantly improved their experience. This, in turn, can help companies retain top talent. Programmers using generative AI-based tools were more than twice as likely to report overall satisfaction and a sense of professional fulfillment. They attributed this to the tools' ability to automate monotonous work that prevented them from performing more satisfying tasks.

Threats in finance

The financial market has always been a target for criminals. With the evolution of technology, illicit financial activities in the digital realm have developed, and were initially called Internet crime, and now mainly are referred to as cybercrime. According to NASK and Cyber Policy[5], the use of AI by criminal groups increases their effectiveness, especially with regard to individual customers in the financial market. This can lead to financial losses, both for the customer and the institutions serving them.

Cyber security and regulation

Artificial intelligence is no longer a technical innovation that is discussed only in academic circles. It is not used only by scientists but is increasingly permeating our daily lives. In order for AI to be employed on a large scale in a reliable and trustworthy manner, it is necessary to discuss the indispensable security measures. In the European Union, work on the AI Act, primarily the regulation of high-risk AI systems, has entered a decisive phase. However, there are some areas that significantly cross the boundaries of the acts. Among these issues is the resolution of what principles we should follow when building AI systems, what unfavorable tendencies may exist in them, and how this technology may be exploited by criminals.

Opportunity or threat?

Technological innovations can inspire equal parts awe and concern. Both of these reactions accompanied us in the fall of 2022, when ChatGPT – a generative artificial intelligence program in the form of a large language model – became widely available, realizing the vision of the unparalleled power of AI. We are all at the beginning of a collective journey toward understanding the power, reach, and possibilities of the technology. In the face of the galloping AI revolution, the need to accelerate digital transformation and retrain the working force is great. Machine learning-based solutions, previously available only to a few individuals, have recently become the preserve of the average internet user. However, it is worth remembering that currently innovative solutions, such as Github Copilot or ChatGTP, are only the beginning of the civilizational and social changes that await us in the future.

The aforementioned tools, as well as the potential of generative AI itself, are already making it possible to optimize and increase the efficiency of the daily activities of specialists in many industries, including IT – programmers, graphic designers, analysts, or security researchers. However, any new technology created and used in the service of humanity, in addition to its undoubted benefits, brings with it a whole range of risks that could affect the way we perceive and use the digital space.

The widespread availability of AI has been recognized and is beginning to be used not only as a tool to streamline work but also to support criminal activities in both disinformation and cybercrime. These threats are not as spectacular as the vision of AI taking over the world in the form of humanoid robots that we know from Hollywood. But the less obvious these consequences are, the more we should keep them in mind.

Sources:

[1] „Sztuczna inteligencja. Czym jest i dlaczego ma znaczenie”, SAS.

[2] „Kurs sztuczna inteligencja dla początkujących. Sieci neuronowe”, Si.

[3] ‘’Generative AI opportunities”, Maven.

[4] Report ‘’The economic potential of generative AI: The next productivity frontier”, McKinsey, 2023.

[5] Report „Cyberbezpieczeństwo AI. AI w cyberbezpieczeństwie”, Nask Cyber Policy, 2023.

Report ‘’The economic potential of generative AI: The next productivity frontier”, McKinsey, 2023.

This article is a translation of a piece published in:

“Kariera w Finansach i Księgowości 2023/24” guide.