Many companies don’t use any AI at all because they have limited (or inexistent) data governance, and there’s the fear of data leaks. Some companies, for example, fear that their data could end up in datasets for training and leak to the public. In 2024, the AI vendors will address the issue with more reassuring licenses, but they may be insufficient for those who classify the risk of leaks as catastrophic.
Generally speaking, new technologies must align with corporate values and user experience, which means that not even the best AI-based tool will be adopted if it disrupts how the company works. That is a problem of personal but also organisational competence. To put it in layman’s terms: a company can’t just switch overnight to AI-based tools revolutionasing its ways of working because the employees won’t follow. It also means that financially capable and well-structured companies are more capable of adopting innovative solutions while smaller firms are not likely early adopters (Na et al., 2023).
Promoting a new technology is always challenging, even when it’s only slightly different from what people are already used to. It’s hard because it requires winning habits and evaluating factors such as training costs and recruitment difficulties.
For example: many companies work in Java and the (relatively) new cool thing in town is Kotlin. Should the companies switch to Kotlin? Many experts love it, and Google promotes it, but:
- Will the current employees be efficient in Kotlin?
- Will the current employees like working with Kotlin if they are used to Java? It’s a change impacting their career.
- Is it easy to find new employees with Kotlin in their skillset?
I never heard of companies developing JVM-based backend services completely switching from Java to Kotlin, Scala or any other alternative language. It may be okay for very small tools with a rarely touched codebase, but, with large single projects, the teams tend to struggle with maintenance since they have to work with a non-standard tech stack.
Considering that a strategic adoption of AI is a huge leap compared to switching between two similar technologies, I am sure it will not happen easily or quickly. Lots of companies have not completed their digital transformation yet! We will see it coming. It will be slow, starting with the introduction of AI in our everyday interaction with technology in our phones and laptops.
For SMEs, the adoption of AI will be slow because, besides the costs, there is a general lack of data governance and legal uncertainty around AI, making this process particularly challenging (Papapostolou et al., 2023). However, initiatives such as the AI-on-demand platform stemming from the European Union’s Horizon program will develop AI-based solutions and support their adoption in the industry (AI-on-demand platform, N.D.)
Today, we are at the beginning of the era of the AI-transformation: we just got ChatGPT, Midjourney, and products incorporating some LLM or ML, but we use them as tools with fine control over the process. In other words, we work with AI just like we did before the AI. At some point in the future, there will be a switching point when the business processes will change to incorporate AI strategically. It will be a massive change, like when we switched from paper to a fully digitalised world.
References
- AI-on-demand platform (N.D.)
- Na, S., Heo, S., Choi, W., Han, S. and Kim, C. (2023) Firm Size and Artificial Intelligence (AI)-Based Technology Adoption: The Role of Corporate Size in South Korean Construction Companies. Buildings, 13(4), p.1066.
- Papapostolou, A. et al. (2023) Encouraging AI Adoption by SMEs: Opportunities and Contributions by the ICT49 Project Cluster.