Artificial intelligence can be a huge opportunity to advance, develop business, or gain a competitive advantage. However, I am also well aware that it is just a technology, and everything depends on how people implement and use it, says Hana Kvartová, director of the Central Europe region at SAS.

Hana Kvartová has held top management positions at SAS's local and regional levels for over ten years. Since January 2020, she has been SAS Regional Director for Central Europe, and before that, she was the Country Leader in Slovakia and the Czech Republic for nearly seven years. She has extensive experience leading diverse international teams and expertise in sales, business consulting, strategic planning, operations management, and communications. Today, she tells us about the opportunities and risks of the increasingly widespread use of artificial intelligence in business and other areas of our lives.

The abbreviation of Artificial Intelligence (AI) has been named the Word of the Year for 2023 by the Collins Dictionary publisher. Will AI still be a top topic in 2024?

Indeed, one does not have to be a high-level tech expert closely following market trends to know what topic resonates most strongly today, not only among IT professionals but also much more broadly, among business representatives or societies. It is artificial intelligence and its wide application in business, the public sector, health care, and many other areas of our lives. It raises high hopes and fears in some, but it is what we are all focusing on today and will deal with this year.

So, what is your approach to this issue?

Calm and balanced. Like any technology, I know artificial intelligence can be a huge opportunity to advance, grow a business, or gain a competitive advantage. However, I am also well aware that it is only a technology, and everything depends on how people implement and use it.

Just a few years ago, regarding digitization, we said that it is not just about implementing the right solutions but is a process that must start with modifying the organization's culture so that digital identity becomes part of it. The same applies to artificial intelligence. It is not a standalone technology that brings benefits on its own. For that to happen, it must be embedded in the organization's strategy that intends to use it.

So, isn't it enough to hook up an AI model to a company's systems so it can access data and have it perform tasks? Of course, I'm making it shallow with such a question, but I'd like you to say more about the challenges of using AI in different organizations.

First of all, we need to know what AI is supposed to do for us and what we expect from it. Once we determine this, we need to train the model in question on selected, representative data to align with our expectations. Here, we touch on an important issue - the quality of the data collected and processed in the organization. Without perfect management and ensuring their high quality, AI will not succeed in any company because the results of its activities directly depend on what data the model has been trained on and what information is then provided to it for processing. If an AI model is trained on biased data, the results of its work may be well-matched.

Imagine we want to entrust artificial intelligence to develop personalized patient treatment plans. A model trained on biased data may prescribe different treatments for the same disease at the same stage depending on the patient's gender, age, economic status, or race. So, it may turn out to be racist, sexist, or apply so-called ageism, and as a result, its recommendations will be harmful to patients. In the financial industry, on the other hand, an AI model trained on inadequate data quality, to which we entrust, for example, the evaluation of a customer's creditworthiness, can firmly shake the company's financial condition if it starts granting loans to people who will not be able to repay their debts. To avoid this type of situation, we must consider data quality.

Moreover, in both examples I presented, AI uses sensitive data. It is, therefore, essential to ensure and adhere to an appropriate regulatory and ethical framework for using AI to avoid abuse and ensure data privacy and security. In addition, it is essential to educate and adequately prepare employees to work with new technologies.

A lot of challenges. Nevertheless, we see that more and more institutions, companies, research units, etc., are leading efforts to develop and implement various artificial intelligence models designed for a wide range of tasks. What is driving them to do so?

Simply put, benefits and competitive advantages. Take the retail industry, for example. Artificial intelligence can enhance responsible marketing, allowing true personalization of sales activities and offers that are fully customizable to customer preferences. By tracking supply chains, AI will also enable shorter delivery times for products to buyers.

AI technologies will also find applications in the insurance industry. They will enable insurers to introduce dynamic premium pricing and risk assessment. They will also help them automate and improve claims processing, fraud and abuse detection, customer service, etc. The banking sector, through the use of AI, will find it easier to counter money laundering, for example, and the process will be cost-optimized.

In health care, artificial intelligence will allow the use of patient avatars to conduct clinical trials and implement individualized treatment plans. The public sector will also reach for it to increase productivity, automate simple tasks, and reduce the impact of a shortage of skilled workers increasingly choosing careers in this area. At the same time, the sector will grapple with the challenge of increasing security for increasingly common digital public services and digital identities.

Finally, the manufacturing industry is where AI is an indispensable part of intelligent factories. According to a 2022 report by the Society of Manufacturing Engineers, their use of predictive maintenance based on AI and machine learning can reduce the number of unexpected failures by as much as 90 percent. This, in turn, translates into a measurable financial benefit - both by reducing the number of costly repairs and, as a result, reducing loss-generating downtime.

All that you say sounds both fascinating and disturbing. On the one hand, the benefits are apparent: AI won't get tired, well-trained AI won't make mistakes, and the results of its work will be fully reproducible and reasonable. However, it comes out that humans are no longer needed.

I strongly oppose the statement that AI will take away our jobs. The steam engine didn't take it away during the first industrial revolution; computers didn't, and we don't have to fear that now. Of course, there will be a change in the labor market - we are already seeing it. We are entering a stage of the emergence of entirely new professions and specializations. At the same time, AI replaces workers in repetitive activities that do not require creativity. This, of course, like any change, raises concerns. However, I know many new jobs will also be created in the above area.

But with the "extinction" of professions and the emergence of others, we will probably have to deal mainly with those organizations where artificial intelligence will be implemented. Where will it not appear? There is nothing to worry about.

It's just that the organizations to which artificial intelligence will not be introduced will be fewer and fewer every year. We must expect a snowball effect here. IT experts are familiar with shadow IT, which involves employees using software on company hardware that has yet to be officially implemented in the organization. Today, the same experts report the phenomenon of shadow AI, which involves using AI models in companies that have not been formally implemented. The source of these phenomena is, among other things, the natural desire of each of us to perform our tasks in the simplest, most efficient way possible, with the optimal commitment of resources. And if some software or AI model offers this, the employee is eager to take advantage of it.

On the other hand, we all must be ready that AI will also be increasingly used by cybercriminals looking to commit fraud, phishing, or hacking into IT infrastructure. Security measures also based on AI models will have to be used to protect against them, which may result in an arms race. All this together, in turn, will make areas unaffected by algorithms, not only in business and health care but also in the public sector, increasingly scarce. We will increasingly come into contact with artificial intelligence daily, and defending against it will resemble a tilt at windmills.

So, will this change that AI brings no longer be stopped by anything?

Observing the current geopolitical situation and considering what we have been experiencing in recent years, I prefer not to see such obstacles arise, as I can only think of factors that could set us back in our development by decades or even centuries. But instead of spinning catastrophic visions, it is better to think about what opportunities artificial intelligence offers us. It can improve and speed up everyone's work, bring competitive advantages to many companies and benefits to their customers, and lead to scientific discoveries and innovations that the world is waiting for. It can improve the quality of our lives, make them more comfortable and more enjoyable, leaving us more time for what we like and what interests us. In turn, it is just another technology - how and to what extent it can be used is up to us.