It's Not the Budget: Digital Culture Makes AI Work

Why do we want to implement AI-enabled systems? Are we ready to master it? Can we guarantee the results by thinking better than artificial intelligence itself? Does our company already think digitally, or does it still reason about its processes in analog terms?

Understanding and building awareness of the foundations for implementing AI-enabled systems is only the first — but most important — step toward a correct implementation. Before acquiring a system or deciding which processes we want to enhance with artificial intelligence, we must first raise awareness of the cultural change that the industrial revolution we are living through represents, and of the new demands it places on the way we think and act, on our processes, and on the culture of our organizations.

Artificial intelligence is a powerful tool for amplifying our capabilities — both positive and negative — but it also amplifies our errors, biases, fallacies, and shortcomings in logic and reasoning. It is not a substitute for our mind and thinking; rather, it is an enabler that allows us to focus our cognitive and reasoning capacity on the important activities where its application is critical, while AI handles those that are repetitive, or that involve analyzing large volumes of data to find patterns, trends, errors, correlations, omissions, similarities, differences, and logical, statistical, and mathematical computations that tend to consume a great deal of time and yield valuable information for decision-making. In terms of the well-known 80:20 rule, AI allows us to focus our minds on that 20% of activities on which 80% of the results depend. 

Five Fundamental Enablers

There are five fundamental enablers of a correct technological transition: digital literacy, the digital maturity of the organization, governance, data integrity by design, and the use of critical thinking throughout the entire process. Enabling our computer systems and processes with artificial intelligence brings many inherent advantages, but it first requires ensuring that people work with, and understand, these enablers.

The human mind and the artificial mind — at least for now — work in similar, but not identical, ways. While the processes of reasoning and analysis may be identical, certain elements set the human mind apart. Current computational memory and processing capabilities are still far from those that biology has granted us. Thousands of years of evolution have allowed us to make better use — though not always for good — of elements such as memory, experience, context, our ethics and moral concepts, intuition, and our heuristics for solving complex problems through the use of intellectual standards.

AI technologies, though highly developed — such as LLMs, with their impressive level of logical reasoning — have not yet reached the speed or the energy efficiency of the human brain. Moreover, they still make errors when trying to fill gaps in information: the so-called hallucinations. 

Similar processes occur in human beings, where we call them biases, fallacies, and prejudices; but our minds find it easier to correct them.

These differences are what make human intervention in the loop a factor of great relevance today: in the end, it is a human being who makes the important decisions and who will assume responsibility for them. It is the filter that guarantees that the results will have, at the very least, the same level of reliability as human reasoning. AI is therefore a tool to make our processes faster and more efficient — not a substitute for our thinking, and far less for our responsibilities. The personnel who control, supervise, and verify AI outputs must have a level of thinking and reasoning at least equal to — or greater than — that of the artificial intelligence they use, and this is achieved through the constant application and practice of critical thinking. 

What This Requires of Us

•       Digital literacy: prepare our people to understand the terminology and rationale of the technologies to be implemented.•

       Digital mindset: have personnel who are comfortable with technology, free of resistance to change, able to think about and visualize processes digitally rather than in analog terms, and able to apply critical thinking to the design, control, and deployment of the technological transition.

•       Digital maturity: ensure that the organization and its processes are already at the appropriate level of technological adoption to move to a higher one, supported by governance and data-integrity policies.

•       Critical thinking: embed and promote its use at every level of the organization for decision-making and problem-solving. 

The Most Common Mistakes

There are many common mistakes when automating processes and, eventually, implementing AI-enabled systems. I will mention the three I consider most relevant:

•       Not knowing what you want: having no clear objective, or taking the step simply because it is fashionable.

•       A lack of process control and definition, which leads to automating disorder.

•       A lack of digital literacy and digital mindset among personnel, together with low digital maturity and the absence of critical thinking throughout the processes. 

A Regulatory Gap — But Not Only That

On another front, we face a significant regulatory gap: we have not yet managed to bridge the digitalization gap, and we are already confronting the new challenge of AI. That said, it is not primarily a matter of regulation either: in the worst case, we can harmonize against more advanced frameworks. The point is that our adoption of these technologies — the way we implement and regulate them — shows serious cultural and process-related lags.

Companies and regulatory agencies alike must make major efforts, at every level, to change the way they work if they wish to remain competitive. What is most concerning is that the gap is widening ever faster.

In this sense, it becomes clear that matters such as budget, regulation, or access to technology are the least of the problems. The main obstacle is neither technological nor budgetary: it is cultural. 

The Real Questions

The question today is no longer whether it is worth using artificial intelligence to optimize and streamline our processes and results. The true, underlying questions are:

•       Why do I want to implement AI?

•       Are my people ready?

•       Does my company have the appropriate level of digital maturity?

•       Do we make decisions using critical thinking?

•       Have we planned this transition properly?

•       And if we fail to do it properly—where does that leave us against our competition? 

Global forecasts on this subject (the OECD, the World Economic Forum, competitiveness indexes, and others) indicate that humanity and its processes will undergo the greatest technological change in history before the year 2030. The largest industrial revolution is already underway, and what will take time is not obtaining the resources to implement or access the new technologies — it will be the mental and cultural shift. After all, that shift is built, it takes time, and it cannot be created by decree.

You and your organization … are you ready?