From Research to Reality: Andrej Svitek on AI in Business

How has the development and use of AI applications/tools for companies changed in the last six months, and what can we expect in the coming months? Not just generative AI, but “analytical” AI as well?

Although artificial intelligence is continuously evolving, the main challenge for companies will be its application in their business operations. According to a survey by Gartner, 70% of companies were in "exploratory mode" a year ago. They were gathering information, especially about generative AI, and thinking about use cases.

Six months ago, most companies were already in "pilot mode," testing various initial solutions with closed groups of users. While current data haven't been published yet, the main challenge companies will face now is how to transition from successful pilot projects into production solutions that can be used across the entire company.

According to a survey by SP CR, companies are most concerned about the misuse of AI for harmful purposes or unauthorized access to data. What should companies focus on to prevent this and ensure AI is beneficial rather than harmful?

We need to focus on what we can actually influence. We can't prevent bad players from using AI to spread disinformation, generate fake content, or create sophisticated attacks. However, each of us can contribute to finding sufficiently beneficial uses for AI so that, at the end of the day, the positives outweigh the negatives. Additionally, we can educate our employees and the general public on how AI works and what pitfalls it entails, thereby minimizing the negative aspects.

There is also a lot of talk about companies experimenting with AI somewhat randomly, without a strategy. Where should companies start, and what initial steps should they take to avoid randomness?

We don't believe companies lack a strategy or are "firing in the dark" with AI. On the contrary, it's admirable how quickly and flexibly even very conservative organizations can implement AI today. What is often mistakenly perceived as "randomness" is the scope and especially the speed of changes that AI implementation brings. For example, a typical IT project (like replacing a data warehouse) might take two years. But the world of AI has changed significantly in just the last two years.

As mentioned earlier, in the past 12 months, most organizations have moved from "thinking about generative AI" to "piloting units to dozens of solutions." Key approaches companies are using include internal AI labs or centers of excellence, educating employees on AI usage, creating central platforms to accelerate pilot projects, and emphasizing the security of solutions.

‍

Author

Andrej Svitek
Head of AI, Trask
Contact: asvitek@thetrask.com