Recipe for the Intelligent Integration of Artificial Intelligence: Secret Ingredients to Ensure the Pie Does Not Become More Expensive Than the Dish It Is Baked In

ChatGPT, Bing AI Chat, Midjourney… It seems we cannot stop talking about artificial intelligence tools. We cannot decide whether we are merely amazed and impressed by what they can do or if we are somewhat frightened by it. Or both. Overnight, there has literally been an explosion of new creativity and experimentation, AI (artificial intelligence) is producing award-winning photographs, songs that become global instant hits, and in addition, it programs, writes journalistic texts, SEO ads, social media statuses, and even business plans. And then we logically ask ourselves, if it can do all this today, when these tools are trained on small amounts of data, what will they be able to do later, when billions of dollars, euros, yuan, and all other currencies from new investors start flowing into the companies that designed these tools. Will AI replace us then? Will it destroy our jobs? Will it, as the tongue bites, turn against us at some point?

Experts repeat that we do not have those answers yet, but that we should not fear catastrophic scenarios. Moreover, they advise that we should not only use these tools now but also leverage them. Especially in business, and to start as an auxiliary tool in organizing people who are experts in their field to save them time and make them more productive.

What You Need

Co-founder and partner at the consulting firm Best Advisory Vedran Antoljak emphasizes that companies operating in Croatia can integrate artificial intelligence (AI) into their business processes in various ways, depending on their industry, size, and needs, and he provided three examples that can also be areas of digital transformation for a business organization. The first is the use of AI within digital marketing activities. The second is the personalization of customer experience and customer support, where we improve customer experience, satisfaction, and loyalty. The third example of how a company can apply AI is in the area of human resources, where with its help we can accelerate management processes, identifying, and selecting talent in the market.

– I believe there is a prejudice that AI tools are expensive and inaccessible. I believe that every small, medium, or large organization in all sectors and industries can and should use AI tools in their business. Namely, many of these tools are free, like ChatGPT, and those that are paid truly justify the subscription cost, which is often very low even for those less financially powerful companies – emphasized Antoljak.

To select the appropriate AI technology for business needs, the first and most important thing, Antoljak emphasizes, is to define business goals, i.e., what we want to achieve by introducing AI into business. At the beginning, it is important to clarify our business needs, goals, and problems we want to solve with AI. The next step that usually follows is technology scouting, a process in which we search for and test available AI tools and platforms that focus on your industry or issues. Then we compare technologies based on functionalities, price, support, and customization needs. Finally, when selecting the appropriate technology, it is important, he advises, to search for case studies and feedback from other users about the selected AI technologies.

Necessary Steps

And when we decide on AI tools, we must not forget about responsibility and transparency, especially in decision-making processes, particularly in sensitive areas such as human resources, finance, and marketing.

– Key measures we can take to ensure the ethical use of AI systems are the development of ethical guidelines, transparency of algorithms, education of employees and partners, regular and independent assessments, involving experts from various disciplines, and continuous communication with users. In addition, in business with artificial intelligence, ensuring data privacy and security is crucial for protecting the interests of clients and the company. To achieve this, it is important to pay attention to data anonymization, i.e., personal identifiers should be removed or masked before processing. In addition to all this, access to sensitive data should be limited to authorized personnel only. The application of data encryption during storage and transmission will further strengthen data protection. Regular security audits and testing are necessary to detect and resolve potential security vulnerabilities. Finally, educating employees about the importance of data privacy and security and best practices in this area will ensure that all team members adhere to the same standards and responsibly use AI technology – explained Antoljak.

Education as an Imperative

He believes that educating employees and stakeholders about ethical and responsible practices in the use of AI in business processes is imperative, just as is educating employees and partners about the opportunities and dangers that AI brings. He recommends that every company that decides to use and implement AI tools organize professionally led interactive and engaging workshops and/or online courses that will attract the attention of employees and other stakeholders on the topic. Consider, he says, also using video content, quizzes, and practical examples to ensure that concepts are clear and interesting. Do not forget to collaborate with internal and external management experts who have experience working with AI technologies, as this will create better content based on real situations (use cases), lessons learned, and best practices.

– This will ensure gaining insights into specific situations where ethical issues may arise, which will help in better understanding the problems. In addition, encourage open communication and the exchange of ideas about ethical and responsible practices in the use of AI among employees. Creating a safe environment for discussion and asking questions can help develop a deeper understanding and better adoption of ethical principles. And finally, do not forget to ensure continuous education for employees. Provide regular training, update content regarding new trends and developments in AI technologies, and encourage employees and stakeholders to continuously improve in this key area, especially in terms of experimentation and trying out – advised Antoljak.

You Are Not an Expert in Everything

Drago Cmuk, partner and director of Mazars Digital, says that integrating any technology into business processes is a strategic issue that involves analyzing processes and needs. In his opinion, if companies identify manual and routine tasks that do not add value and if they also conduct an analysis of the cost of manual labor, then it becomes very easy to decide. However, there is another problem.

– Often, the capabilities of technology are overestimated, leading to situations where the pie is more expensive than the dish. Every industry has areas that can be relatively easily identified and automated if approached systematically. Usually, it is best for external experts to handle this. Robotic process automation is one such area where any clicking or retyping from paper or from application to application is usually automated. This is resolved very quickly and efficiently today. But there are also many other, more advanced applications, and my advice is always the same: start as soon as possible with small, brave, and wise steps – emphasized Cmuk.

When asked how to choose the appropriate AI technology for business needs and where to find experts for implementing AI into business processes, Cmuk responds that there are already many technologies available and new ones are being developed every day, and the best advice he can give people is to stop trying to be experts in everything.

– People should focus on jobs where they are experts and where they bring the highest value. It is probably not the smartest to fix their own car, plumbing, or electrical installations. Likewise, there are people who deal with these technologies, and they should be consulted, especially consultants who know specific industries, which is always the fastest and most cost-effective, and the answers are unbiased or sugar-coated – responds Cmuk.

And how to do all this responsibly is another matter. AI cannot be responsible because it lacks consciousness; it is always the people who use and/or program that technology who are responsible. Furthermore, do not forget, Cmuk reminds, that artificial intelligence does not exist, meaning it is not about intelligence but about very powerful algorithms that operate according to more or less predetermined rules.

AI Literacy

Aco Momčilović, co-founder of the Global AI Ethics Institute, says that in this whole story, it is first important to educate managers and entrepreneurs about what AI actually is, what AI tools exist, to promote AI literacy, and only then to continue the discussion in the direction of what it can or cannot change in companies. That there is interest in such education is confirmed by the fact that he receives a large number of inquiries for conducting in-house training on the basics of AI and some of the technologies in that spectrum.

– First, we need to educate and inspire. After that, it is often very useful to hold workshops where employees can think about which processes they can incorporate existing AI tools into. And after that, implementation comes next. These do not have to be big steps and strategic things, but everyday processes that can be improved very quickly – emphasized Momčilović.

Considering that he also deals with ethics in AI, he says it is important for companies to have clearly defined ethical guidelines for using AI systems in sensitive areas. They should ensure that these systems are used fairly and justly towards all stakeholders, and they should also contain plans for correcting errors.

– First, it is important to ensure that only quality and reliable datasets that comply with data privacy regulations are used. These data should only be collected with the user’s consent, and users should be informed about the reasons for data collection. Furthermore, it should be ensured that reliable and secure AI technologies and tools that comply with the latest standards for data security are used, but we will only address that when we all learn what AI is and how it can help us – concluded Momčilović.