Mornings in the offices of chief technology officers now resemble the siege of a fortress, with headlines about breakthrough language models constantly hitting the walls. Hovering over the desks is the question that falls from the mouths of CEOs with the frequency of a mantra: “Why don’t we have this yet?”.
This phenomenon, aptly christened the ‘AI moment’ by Mark Baker, has introduced a specific kind of nervousness into corporate corridors. The line between visionary and panic management has become dangerously blurred. The Altimetrik report casts a chilling light on this situation, revealing that most organisations have thrown themselves into the deep end without checking whether they can even swim in the new regulatory and operational environment.
Architecture in a hurry and foundations made of sand
Statistics can be ruthless for enthusiasm without a plan. Only 14% of companies implementing AI-based solutions have a clear strategy that goes beyond general declarations of innovation. 71% are operating in a state of permanent construction, where foundations are being poured at the same time as decorative turrets are already being installed on the roof. This lack of embedding in specific business objectives means that, instead of becoming an engine for growth, artificial intelligence becomes a technological debt incurred at an extremely high interest rate.
The responsibility for this state of affairs has traditionally been pushed onto IT leaders. They find themselves caught in a vise of expectations: on the one hand, the pressure for immediate results, on the other, the lack of management systems, training frameworks or clearly defined crisis paths.
Implementing tools before guardrails are established is akin to trying to control a nuclear reactor with a toaster instruction manual.
The paradigm of uncertainty in a deterministic world
Business has relied on predictability for decades. Traditional IT systems were deterministic: specific inputs always resulted in an identical outcome, and algorithms strictly adhered to predefined rules. In this world, it was easy to pinpoint the culprit of a failure or process error. The advent of probabilistic systems such as generative models has turned this order upside down. AI does not operate on certainty, but on probability.
This transition requires a new form of competence from technology leaders – uncertainty management. Since the outcome of a system may be different every time, legacy operational procedures become useless. Building accountability in such an environment requires a deeper commitment to testing and planning than was the case in any previous wave of digitalisation.
The AI moment forces the question of who will pick up the phone when the algorithm, in a fit of statistical hallucination, makes the wrong financial or image decision.
The trap of an accounting view of innovation
One of the most worrying signs coming from the market is the motivation behind the adoption of artificial intelligence. Most decision-makers point to the reduction of operational costs as the main objective. This approach, however, is confusing effect with cause. Savings are the fruit of a well-designed strategy, not its foundation.
Attempting to implement AI solely under the dictates of a spreadsheet leads to superficial implementations that, in the long term, generate additional expenses related to bug fixing and lack of scalability.
Understanding the return on investment (ROI) for artificial intelligence is based on the same mechanism as always: identifying the problem precisely, developing an adequate solution and meticulously calculating the savings from its application.
Jumping straight into the pilot phase, without asking the right questions about the goal, is a strategic mistake that Mark Baker calls ‘having a solution and looking for a problem for it’.
Human capital on starvation rations
Technology, regardless of its degree of autonomy, remains anchored in human action. Meanwhile, the data on employee AI education is alarming. Nearly eighty per cent of respondents admit that their teams receive less than ten hours of training per year. This is a glaring mismatch between investment in software and investment in the people who are supposed to operate it.
The result of this gap is growing uncertainty. Almost half of managers and employees feel left behind, which breeds a natural resistance to change. Instead of a proactive retraining strategy, many companies opt for a wait-and-see strategy, hoping that job roles will be reduced through the natural expiry of positions.
This is a passive approach that wastes the potential of human-machine collaboration. Mature AI implementations are those that have invested in trust and behaviour modification, not just API access.
Breathing strategy as the latest technology
Faced with the ‘frightening speed’ with which AI is encroaching on businesses, the most innovative leadership move may be to paradoxically slow down. Taking a deep breath and getting back to the basics of technology management helps to sift the hype from the real business value. Managing the AI moment is not about being first in line for every innovation, but about building a structure that can withstand the weight of the new reality.
