You are here:

AI requires skills: Data literacy as a prerequisite for the mature adoption of artificial intelligence

AI REQUIRES SKILLS: DATA LITERACY AS A PREREQUISITE FOR THE MATURE ADOPTION OF ARTIFICIAL INTELLIGENCE

By 2026, artificial intelligence will cease to be merely a technological innovation and will become a key component of business and digital infrastructure. Organisations are moving from experimenting with AI to its widespread implementation in operational, decision-making and product processes.

With this transformation, data becomes a strategic foundation – its quality, management and interpretability determine competitive advantage and the effectiveness of AI systems. Based on this data, our AI assistant provides its response to our query. And this, unfortunately, is where we reach the point where not everyone knows how to handle the information they receive from their AI assistant. They assume this information is correct, without any deeper analysis or reflection. Until, at a business or social gathering, one of our conversation partners points out that the information is false, fabricated or – to use technical terminology – a hallucination of the AI model.    

AI IS BASED ON PROBABILITY, NOT FACTS

Generative AI models are based on vast amounts of data and probability. They analyse our query and generate a response based on the probability of a particular combination of words appearing within a given subject area.  

The developers of these models strive to make this probability as high as possible, but it will never be 100 per cent. That is why errors – or, as I mentioned, ‘hallucinations’ – occur in the details.  

A key aspect of using AI tools is therefore the ability to critically evaluate the responses obtained. This competence requires the ability to:

  • read the information,
  • understand its context,
  • interpret its meaning,
  • drawing appropriate conclusions.  

DATA LITERACY AS AN INFRASTRUCTURAL COMPETENCE

The ability to understand and interpret data (known as ‘data literacy’) is no longer the preserve of analysts confined to BI departments. It is becoming a universal competence – the foundation of everyday life in a world where almost every decision has a numerical basis. It is no longer just an ‘added-value’ skill. It is a prerequisite for informed action – both professionally and privately.

The modern user cannot afford to ignore data. They are surrounded by it and – whether they like it or not – must grapple with it every day. Data:

  • financial,
  • health,
  • consumer,
  • organisational,
  • sales

constantly influence our choices. They show us how much we earn and how much we spend. They provide information on the quality of our sleep and our physical condition. They suggest which products customers are buying and which areas of the business are growing faster than others.

We can ignore them. We can also unthinkingly accept the first interpretation – often generated by an algorithm. However, it is only through conscious analysis that they gain value. In each of these areas, the ability to interpret data correctly – to understand the context, proportions, limitations and consequences of the decisions taken – is of key importance.

In an era where AI can generate answers in a matter of seconds, it is precisely this human ability to understand data that acts as a stabilising force within the system. Without it, technology remains fast but not necessarily accurate. With it, it becomes a genuine aid in making responsible decisions.

EXAMPLES OF DATA LITERACY IN PRACTICE

Let’s look at our finances. We are in the tax return period. By analysing our PIT return, we are able to prepare a summary of our past year, such as our average revenue or income. It is worth bearing in mind the possibility of claiming tax reliefs, which also requires us to go through the process of analysing the available options. By understanding which reliefs are suitable for us and drawing the right conclusions, we can fill in the relevant sections of our tax return. We have to do this ourselves; the system won’t do it for us. It might try, but errors will usually occur. Of course, an accountant can do this if we use their services. However, it is always worth checking how the return has been prepared.

Staying on the subject of finances, we can take it a step further. If we have an income, we also have a household budget that needs to be managed in some way. Analysing what we spend our hard-earned money on – bills, rent, loans, groceries and other items – will help us plan our expenditure properly. When analysing our expenditure, it is worth bearing a few aspects in mind. Comparing one month to another can be misleading. Firstly, the months of the year have different numbers of days. Because of this, in a given month – for example, February – we may spend less on our needs, whilst in another month we may spend more. Another factor is public holidays and time off. During these periods, we tend to spend more, which can also be planned for accordingly. For example, what we manage to save in February – due to there being fewer days in the month – we can spend on a May holiday or some other event. Analysing our household budget will also help us protect ourselves against unexpected situations that can happen to any of us, so dedicating time to this will also have an impact on our financial security.

Another example relates to what we wear on our left or right wrist. It’s a smart band, a smartwatch or another device that monitors our daily activities. It tracks the number of steps we take, our heart rate, oxygen saturation levels, sleep quality and many other things, which we can then analyse using a mobile app to which our smart device is connected. How many of you reading this have actually taken the time to analyse the data collected and draw conclusions from it? To make changes – where possible, of course – in order to achieve better health outcomes. After all, we’re doing this for our health and longevity. The problem is that the app doesn’t provide information on what the appropriate metrics are for a specific user, as there are many factors that need to be taken into account. So in this case, we either need to consult external sources or speak to a suitable professional, such as a doctor, to determine whether the parameters being measured are appropriate. It’s worth doing this for your own sake.

We must therefore bear in mind that the mere availability of tools does not guarantee value. The effective use of these tools depends on the users’ level of competence in understanding and interpreting data (data literacy).  

THE EUROPEAN CONTEXT

According to Eurostat data, around 85 per cent of Europeans living in larger cities possess basic skills in reading and understanding the data presented. In smaller towns, the figure is around 80 per cent, and in rural areas 78 per cent. Unfortunately, in Poland, the differences between urban areas are greater: 88 per cent in larger cities, 83 per cent in smaller towns and 77 per cent in rural areas. On the other hand, there are countries in Europe, such as the Netherlands, Iceland and Norway, where there are virtually no differences between urban areas and where data literacy rates are close to 96–98 per cent. This suggests that, despite Poland’s good results, there is room for improvement and for raising public awareness of the need to understand what we read, how we analyse charts and what conclusions we should draw.

CONCLUSIONS

In summary, the full implementation of AI tools – both in professional and private life – requires more than just access to the technology. It requires maturity in reading, understanding and verifying the information we receive.

Entrusting the responsibility for assessing the accuracy of data solely to an algorithm can lead to undesirable consequences – ranging from minor operational errors to serious decisions based on imprecise or erroneous premises. Therefore, the results generated by AI tools should always be cross-checked against the user’s knowledge, experience and – where necessary – additional sources.

It is precisely this moment of verification that distinguishes the unthinking use of technology from its conscious application.

Through critical analysis, the decisions we make become more accurate and better grounded in actual data, rather than in its seemingly convincing interpretation. As a result, the efficiency of our work, the quality of our actions and the soundness of our decisions improve – whether they concern everyday choices or key business matters.

AI can significantly speed up the decision-making process. However, responsibility for the quality of those decisions still lies with humans.