Executive Surveys and Their Connection to Reality (Category AI)

While taking a closer look at the report by Yakov and Partners and Yandex, I decided to examine its sources of information. There are mainly three: surveys of company executives, vendor representatives, and users. Reading executives’ accounts of AI adoption reminded me of Jaroslav Hašek and his Good Soldier Švejk. In particular, statements like this:

According to the CTO survey, generative AI has moved far beyond isolated experiments in two years: the average number of functions with pilots or full deployments rose from 2.4 in 2023 to 3.1 in 2025, and the technology is now used in 80% of core business functions.

It brought to mind Hašek’s portrayal of ministers, generals, and senior officials seemingly suffering from a collective madness that inevitably spreads down the hierarchy. Information from above becomes orders, directives, and propaganda; further down, it turns into absurd initiatives, harsh punishments, and junior officers and soldiers rushing around nervously. What travels back up is not orders, but denunciations, distorted rumors, and statistics of “success” polished to meet the bosses’ expectations. That creates an illusion of order and rationality.

In that respect, I place more trust in MIT’s The GenAI Divide: State of AI in Business 2025. It reports that despite global investment of $30–40 billion, about 95% of organizations saw no measurable return from GenAI projects, while the remaining 5% managed to make money from them. The authors attribute this to different approaches to adoption, rather than model quality or regulation. I covered the MIT report in more detail here. After the weekend, I will return to the report by Yakov and Partners and Yandex.

#Humor #AI #Engineering #Management #Leadership #Economics