Management Blog and case study

Whether we’re talking about baking bread, self-diagnosing medical conditions, constructing a building, or even figuring out if we’re feeling okay today, the standard response has become: “Have you asked the AI?”

This slightly sarcastic introduction is not intended to disparage or downplay the scale of the phenomenon—a major technological trend with societal impacts whose full extent we cannot even yet assess.

The goal of this post is to shed light on the opportunities, but also the risks, that AI may present when implemented in a business setting.

But let’s start at the beginning, as they say, and place the technology in its broader context.

 

“How It Works”

From the perspective of its application, there is no single form of AI, just as there is no single field of knowledge or type of problem to solve in science.

To put it simply, AI is an engine, an analytical model, resulting from the integration of three components known as “analysis rules” (“Context”), rules applied to a dataset (“Content”) to answer a user’s (“User”) questions via a query interface.

This analysis engine is driven by a set of mathematical rules that replicate or apply a logic associated with a form of human intelligence, but which remains in the raw form of “0”s and “1”s. Two schools of technical thought frequently clash—the logicist approach and the neural approach—but the goal remains the same: to answer a question as best as possible, just as a human would.

These may include AI models specialized in medical image recognition and analysis, conversational agents, artistic creation, etc., and from this you will understand that the engine’s algorithms and mathematical rules can vary significantly.

For example, in the case of the most well-known conversational agents such as “ChatGPT (OpenAI),” “Gemini (Google),” or “Copilot (Microsoft),” it is known that their analysis engine relies heavily on a probabilistic logic of the “best next word” to add to the sequence of the response to your query.

In some cases, and especially looking ahead to the future, AI models exhibit characteristics of continuous improvement—whether in their reference data or their analysis rules—through a process of continuous learning and feedback (“Deep Learning”). This is the key to a true embryo of intelligence, since it is the process by which the machine is able to continuously improve itself—apparently, on its own.

We draw your attention to our phrase “continuously improve itself—apparently, on its own.” And even more specifically to the phrase “apparently.”

Because what we wish to highlight here is that we, as human beings, have nurtured and continue to nurture the beast, which is becoming less and less… beastly! This is an observation we can make at the current stage of artificial intelligence technology, as eventually these systems will very likely be able to generate their own new knowledge. There are already examples of this.

This observation is important for the rest of the article, as it relates to the risks of unintentionally publishing your company’s sensitive information into the public domain. We’ll return to this at the end of the article.

At this stage, simply note, for example, that the intelligence of AI chatbots is solely and directly linked to the intelligence of all its users—the creators of the reference data—and thus to their way of “thinking.”

If you ask a far-fetched question or use extremely poor syntax, don’t be surprised to get a far-fetched answer or one that’s completely off-topic.

The “Force” and its dark side

As with many things, AI has two sides to the coin. Extremely positive potential benefits:

• In the medical field, think of faster diagnoses and the discovery of new drugs or treatments for diseases currently deemed incurable. Our quality of life will benefit from this.
  
• On the environmental front, think of discoveries in new manufacturing technologies with very low greenhouse gas emissions, innovative waste treatment methods, new soil revitalization strategies, or, more broadly, the optimization of circular economy processes to ensure the planet is fed equitably.
  

On the flip side, there will always be profiteers and those with malicious intent who will exploit AI’s impressive capabilities to take advantage of the vulnerabilities of society, business, and human beings at every level.

The most concerning issue on the horizon is the potential gradual erosion of human critical thinking—that is, the ability to think for oneself—and, by extension, the end of humanity’s supremacy over the capacity for sound judgment. Establishing an ethical framework for the development of such technologies is therefore becoming increasingly essential.

Evolution or revolution?

Given all the media coverage of the phenomenon, it makes sense to view AI as an exceptional innovation and novelty.

However, the commercial beginnings of AI date back to the early 1990s, with the media impact of the famous defeat of chess champion Garry Kasparov by IBM’s “Deep Blue” computer in 1997 being particularly noteworthy.

The field was officially named in the 1950s, with the term coined by researcher “John McCarthy” at the Dartmouth Conference, and the renowned mathematician “Alan Turing” had identified the field as early as the 1940s. Regarding the turbulent history of AI, it’s worth noting that global research projects on the subject were even abandoned twice in the ’70s and ’80s.

Without getting bogged down in technical details, we can say that the Google search engine is at the dawn of today’s democratized AI. We also note Google’s rapid integration of its “Gemini” AI engine into its well-known search engine, as it is now the one providing the initial results for our Google searches.

In short, the concepts and principles surrounding Artificial Intelligence are therefore not new. What is new is the phenomenon of the accelerated democratization of a technological potential:

“An intricate amalgam of technologies converging at a critical juncture, combined with enhanced technical capabilities and power that surpass a threshold—a software and hardware ‘trigger’—to create a disruptive technological phenomenon capable of redefining, yes indeed, the way we think and therefore live. ”

That said, the question facing this new disruptive technology:

Is all the knowledge and expertise we’ve accumulated within our company no longer valid? Does this accumulation of corporate knowledge no longer represent a solid foundation on which to base our management decisions?

To ask the question is, of course, to answer it:

No, AI offers exceptional possibilities, but AI must be viewed as a tool for evolution within your company and not as a trigger for a complete redefinition of your corporate “DNA.”

From Technology to Smart Management

Ever since humans have exercised their creativity and invented new tools, their motivation has been driven by a desire to do more quickly and better with fewer resources, time, and/or energy—and thus with less effort:

“The invention of the wheel, the invention of writing, the invention of the printing press, the invention of the steam engine, the invention of electricity, the radio, the telegraph, and the telephone, the invention of the automobile, airplanes, the invention of computing and the internet, etc.”

Steve Jobs, founder of Apple, could not have put it better when he spoke of his personal computer project, which he described as “a bicycle for the mind”—a tool to help human thought go further and faster.

Consequently, simply based on the fact that any dynamic company actively seeks to grow faster and in the most optimal way possible—that is, by maximizing its profitability—the question is not “Why AI?” but rather “How?” and “When?” you will leverage AI in your business.

Here are three possible areas for integrating AI into your business:

Automation

In this first area of focus, think of AI as a lever for improving various operational aspects, for example:

• Accelerate your development or the enhancement of products or new services
• Improve the quality of products or services
• Increasing the speed of product or service delivery
• Reducing product manufacturing or service delivery costs
• Amplifying your promotional efforts, accelerating them, and expanding their reach

At the heart of all these levers, consider the concept of “process automation”—that is, accelerating routine processes to improve productivity and quality.

Decision-Making

In a second area of focus, AI can be an effective ally for accelerating management and business analyses. For example:

• Quickly provide a summary of the economic context of your industry based on key indicators;
  
• Quickly provide an accurate summary of large volumes of information
  
• Provide a well-supported financial analysis and forecast.
  

At the heart of this type of AI application, consider the concept of a “decision-support” tool, viewing AI as a platform for tactical and strategic thinking.

Innovation

In a third area of focus, consider using AI to discover new business avenues for your company. Extensions and additions to your business model under the theme of “diversification potential” for your activities in various possible forms.

At the heart of this type of AI application, consider the concept of “Ideation”—AI as a platform capable of helping you discover new opportunities for your business.

For a concrete example of the strategic leverage of this latter use of AI, it is well-documented in business literature that several major corporate successes have been achieved by applying proprietary technology in a business sector other than their original vertical. And the reverse is also documented: companies have discovered an exceptional lever for differentiation from their competitors in a specialized technology used by a sector completely different from their own.

Recommendations on the Use of AI

While it is clear that AI can no longer be ignored, the final step is to establish the parameters of its framework for use—that is, to answer the question, “How should AI be leveraged?”

As we have seen, AI presents not only external challenges—such as malicious actors who could use AI to target you personally or your business—but also internal management challenges and pitfalls. Here are a few examples:

• The unintentional disclosure of sensitive and confidential information, whether it pertains to customers or employees. Consider the potential for legal action resulting from unintentional violations of applicable privacy laws.
  
• The publication of strategic information about your company in the public domain. Think of your trade secrets or intellectual property.
  
• The lack of a reflex to step back, analyze, and verify sources regarding the results and responses proposed by AI. Consider the possibility of overly hasty, ineffective decision-making that could have major negative impacts on the company.
  

It is therefore essential to establish basic mechanisms to govern the use of AI to ensure that your use of the technology always generates added value and does not create problems for your company.

Let’s focus on the unintended and problematic dissemination of sensitive or strategic information into the public domain to highlight an important nuance regarding AI deployment.

Internal Encapsulation of AI

Many of the AI solutions available on the market offer options for deploying their analysis and search engines in a way that limits their use strictly to an internal network—meaning your private corporate information is never shared externally, and your strategic information is never publicly disclosed.

When this type of deployment is possible, you already benefit from a form of native, de facto protection for your day-to-day use of AI. All your users operate within and are constrained by a basic technical control framework.

This type of deployment should, of course, also remain subject to the additional recommendations in the next section, but internal encapsulation already provides a more solid foundation.

Policies for Expanded AI Use

The reality is that you will not be able to control all the scattered uses your employees might make of AI on a daily basis—for example, from their personal phones—without your knowledge or any record of it. This is where a more formal and robust corporate framework for use must be developed.

This minimum framework should take the form of written policies in the form of an internal guide or standard, published to all your employees. In this AI usage guide, it is recommended to include clear answers and guidelines on the following aspects:

• “Who” has access and can use and/or authorize the use of AI for business purposes?
  
• What types of queries and questions can be submitted to the AI? This includes specific details such as “What syntax should be used?”, “What words should not be used?”, “What private information must never be included in a query?”, etc.
  
• A protocol for authorizing and double-checking responses obtained from the AI before taking any resulting action.
  

ERP deployment and field applications

The preceding discussion presents AI from a largely strategic, if not tactical, perspective, but the application of AI in a manager’s day-to-day work can also be very concrete and hands-on.

For example, in the context of an ERP solution deployment, AI can be a highly useful tool to support and automate tasks such as the following: All of these possibilities must always be approached, once again, with the understanding that AI does not replace the functional analysis or quality control performed by your internal experts and your AI business partner. They must remain at the center of the validation and final confirmation of deliverables generated by AI.

• Create test databases;
• Analyze, convert, and export/import historical data;
• Anonymize sensitive data;
• Automate the creation of internal procedure manuals based on “live” usage videos;
• Automate the management of a continuing education program on key topics for training existing staff and onboarding new hires.

 

Conclusion

The scope of AI’s potential is exceptional—and as vast as the pitfalls its use may entail. In conclusion, there is only one central message to convey to you in the form of a general recommendation:

You are and remain the “captain” of your company. Your business vision, your expertise, and your business knowledge remain the cornerstone of your success.

To echo the first post published on our blog:

Artificial Intelligence (AI) must be viewed and approached solely as “Dont' confuse Tool and Solution”

One thing is certain: by using AI wisely, some companies will be able to accelerate their growth and stay ahead of the competition. And companies that don’t believe in it—that fail to give it even minimal attention—may eventually have to close their doors sooner than expected, either because their business model becomes obsolete or because of more effective competition.

Good luck with your management!