It is no exaggeration to say that the democratization of new forms of artificial intelligence (AI) such as ChatGPT (OpenAI), Gemini/Bard (Google), and Copilot (Microsoft) is a social revolution in the digital age.
The mainstream use of AI systems will have a disruptive impact on many sectors, including university education, the legal system, and of course the world of work.
These changes are happening at an astounding pace, and research is struggling to keep up. For example, in just a few months, the ChatGPT platform has improved so much that it is now capable of placing in the top 10% of highest scores on the U.S. Uniform Bar Exam. In response, some U.S. law firms are replacing paralegals with AI software that detects judge preferences to personalize and automate pleadings.
But while the technology is making impressive advances, the promise of AI doesn't line up with what we've learned from more than 40 years of research in organizational psychology. Drawing from my years of experience as a strategic management professional, I'll shed a clear, yet complementary, light on the dark side of organizations — their irrational (or foolish) behaviors and procedures — and consider how these will impact us when AI is added to the package.
Stupid organization
Have you ever found yourself in a professional situation where your idea was more creative and less costly, but was rejected with the answer “rules are rules”? Congratulations! Science tells you that you worked (or still work) for a stupid organization.
Organizational stupidity is inherent in all organizations, to varying degrees. It is the result of human interactions that de factoinefficiencies, and the processes that control work (such as company policies) risk making the organization itself stupid unless they are regularly updated.
Some organizations work hard to update themselves, but others, lacking the time or in the pursuit of everyday convenience, maintain processes that no longer fit the realities they face. And they become stupid. Organizational stupidity has two components: functional stupidity and organizational incompetence.
Functional stupidity
Functional stupidity occurs when the actions of an organization's managers impose discipline that limits employee relationships, creativity, and self-reflection. In such organizations, managers reject rational reasoning and new ideas and resist change, which leads to increased organizational stupidity.
As a result, employees avoid working as teams and spend their professional resources (knowledge, expertise, etc.) on personal gain rather than the organization's benefit. For example, an employee may notice warning signs of a machine breaking down at work but decide to say nothing because “that's not my job” or because they believe their boss will be more grateful for fixing the machine than for preventing it from breaking down.
In the context of functional stupidity, integrating AI into the workplace only exacerbates this situation: Employees with limited connections to their colleagues and seeking to accumulate as many professional resources (knowledge, expertise, etc.) as possible tend to increase information requests to the AI. These requests are often made without contextualizing the results or without the necessary expertise to analyze them.
For example, organizations plagued by functional stupidity traditionally tasked employees with analyzing market trends and then handed that information off to another team to prepare advertising campaigns. The integration of AI risks everyone in the organization (whether they have the expertise to contextualize the AI's responses or not) looking for new market trends and trying to predict them. of Best idea in a meeting in front of your boss.
Examples of functional stupidity have already been in the news: for example, in one court case, a US law firm (with the help of ChatGPT) cited six completely non-existent precedents. After all, this kind of behavior reduces the efficiency of an organization.
Incompetent organization
Organizational incompetence stems from the company's structure: it is the rules (often inappropriate or too strict) that prevent the organization from learning from its environment, its failures and its successes.
Imagine you're given a task to complete at work. You have one hour to complete it, but the deadline is set for the end of the day. Because completing the task faster doesn't provide you with any benefits, such as additional tasks to complete or a reward for working faster, you may be tempted to stretch out the time you need to complete the task to the limit. As a result, you're putting the Parkinson's Principle into practice.
In other words, work (and the cognitive load required to perform it) is adjusted to meet the deadline. Organizations that are prone to Parkinson's principle are unclear to what extent the use of AI can improve work efficiency.
The second element of organizational incompetence related to integrating AI into the workplace is the principle of “kakistocracy” – that is, the individuals who seem least capable of holding managerial positions end up in them.
This situation occurs when an organization decides to promote an employee based on their current performance, rather than on their ability to meet the requirements of the new role. This causes promotion to stop the day an employee is no longer competent in the role they currently hold. If all promotions in an organization are done this way, it creates a hierarchy of incompetent people. This is known as the Peter Principle.
Peter Principle is even more pernicious in organizations that integrate AI. For example, an employee who can write programming code in record time to solve some time-consuming problem at work and master AI faster than his/her peers will have an advantage over them. This skill will give them an edge in performance reviews and may even lead to promotions.
Incompetence and inefficiency
But employees’ AI expertise alone will not be enough to handle the conflict resolution and leadership challenges that new management brings. If the new manager does not have the necessary interpersonal skills (as is often the case), they will likely suffer from “ingeritance” (a combination of incompetence and jealousy) when faced with these new challenges.
This is because the new manager will not be effective if they reach the limits of AI when human capabilities (creative thinking, emotional aspects of relationships) need to come to the fore. Managers who feel incompetent tend to take more time to make decisions and try to find solutions to non-existent problems in order to show off their technical skills and justify their expertise to the organization. For example, the new manager may decide that it is essential to monitor (with AI, of course) how many keys the employees on her team type per minute. Of course, this is not an indicator of good performance at work.
In other words, it is a mistake to think that a rational tool like AI in a non-rational environment like an organization will automatically result in the efficiency gains that managers expect.Above all, managers need to make sure their organization is not stupid (both in terms of process and behavior) before considering integrating AI.
