Artificial intelligence is changing companies from the inside: it holds up a mirror to their culture, processes, and management readiness. At the same time, it reveals who has the appetite to innovate – based on real-world experience, about a third of people already use it routinely. Lukáš Benzl from the Czech Association for Artificial Intelligence talks about where AI delivers real value and where it runs into obstacles.
AI‑first approach in practice
AI‑first companies build on the premise that most problems will be solved with the help of artificial intelligence across departments. It's not just for startups; an established company seeking an impetus for growth can also adopt an "AI‑first" mode. The key is a willingness to test, to tolerate dead ends, and not to punish employees for experiments the company learns from.
An example is the online supermarket Rohlík, which uses AI in expansion, logistics, and in conjunction with robotics. For instance, it dramatically reduced the cost of producing a commercial: whereas last year it cost 390 000 €, this year thanks to AI only about 400 €. Steps like these show how AI can streamline the business and move the company ahead of the competition. But a clear objective is essential – not "AI for AI", but for specific value.
How to start and where AI makes sense
It's worth first setting internal guidelines and paying for quality tools for people – the investment of tens of euros per month usually pays off. In practice, deployment can be split into personal productivity (licenses in the Microsoft/Google ecosystems, training, internal "AI ambassadors") and "core" projects. For the latter, you need order in data and processes; otherwise you risk wasted money and disappointed expectations. According to a recent MIT study, the vast majority of generative AI projects fail precisely because of poor expectations and companies' lack of preparedness.
For small and medium-sized companies, AI saves costs and increases capacity, especially in accounting, marketing, or development; it also makes mapping foreign markets and expansion easier. The impacts are also felt in the labor market: demand for juniors is declining and in customer support AI already handles 80–90 % of inquiries. Instead of blind faith in "new professions," the decisive factor will be the ability to learn quickly and move juniors into more senior roles directly within companies. You don't always need your own model – often a smaller, locally runnable one (e.g., Lama) is enough, which also brings security advantages; alongside that, it's important to cultivate critical thinking and not be ashamed of using AI.
