Bringing The Managerial Vision Closer To Artificial Intelligence

Artificial Intelligence

In the last two years, the avalanche of news around Artificial Intelligence (AI) has been overwhelming for everyone. The rate of appearance of new tools is so high that it is difficult to decide which one can really add value to our work, so that we decide to invest time in testing it to integrate it into our daily lives and into our teams.

This is especially critical in the case of management profiles, who have the responsibility of promoting innovation in our organizations from our position as agents of change.

If we spend time trying out any new tool, we can fall into a cycle of losing effectiveness and focus.

Without being dazzled by all this window dressing, professionals have to learn to identify what can really add value in the short and medium term because, in the long term, we don’t even have time to consider it before it becomes obsolete.

Although AI has been with us for a long time, it has been in the last two years when there has been an explosion at the user level, with the popularization of ChatGPT, which has triggered the race among all the actors involved in the development and commercialization of software.

We consider these tools “smart” because they are capable of learning.

Before entering into the tools themselves, I am going to take a moment to explain the basic structure of AI and distinguish the two large areas that we can currently exploit in AI at the user level.

Broadly speaking, AI is based on a triangle formed by three key elements:

Data is the knowledge that we make available to the system.

Software is the intelligence that takes advantage of knowledge, generating value for the user.

Hardware: is the force capable of applying intelligence to knowledge.

The explosion in recent years has been possible due to the increase in the computing capacity of the new

systems developed. However, this capacity requires enormous consumption of resources that puts sustainability in check.

When it comes to distinguishing the basic types, we have two large areas:

On the one hand, we have Analytical AI, which has been with us for longer and allows us to clarify and analyze data, with predictive analysis capacity.

Secondly, more explosive, although it was the last to arrive, is Generative AI. It is capable of creating new content from data with which the models have been trained.

Analytical Ai

Within these sections we have different solutions:

Sentiment analysis. Based on the repetition of certain words that give us a positive, negative, or neutral tone.

Reputation analysis. On the keywords related to our brand.

Recommendation. Using algorithms based on our use of content platforms, the system is capable of recommending new content (Netflix, Spotify, TikTok, Google).

Stylometry and content quality. Through content patterns it helps us make our communication clearer.

You can feed the tool with our organization’s style manual and then upload any text so that it can give us recommendations for improvement and confirm that it matches the company’s style manual.

Creation of thematic maps. Through a selection of text sources, news and conversations on certain platforms, it is able to identify the narrative by finding patterns that represent thematic clusters.

Business intelligence. Converts information into knowledge to facilitate management decision-making (Dashboards, Reporting).

Prediction. From historical data, it provides a probability range. It is used in scenario analysis and data mining.

In all these cases, AI provides results with uncertainty that must be interpreted correctly.

Generative Ai

It has been the great revolution of recent years. It is capable of creating new content from patterns learned in existing data. We can use it as an extension of our creativity.

One of the most important changes is that it is no longer necessary to know how to program to obtain results. The appearance of the prompt allows us to interact as in a conversation.

Generative AI is incorporating some features of Analytical AI so that we can ask questions through the prompt and obtain analytical data.

Within the evolution of Generative AI, there have been two consecutive contrasting movements: Divergence vs. Convergence.

Divergence has led us to create many options, making a multitude of tools available to the end user.
Convergence leads us to make decisions and choose a path to take advantage of these tools.

Although there are many solutions, the vast majority are based on OpenAI’s GPT-4 model. The two main references are ChatGPT and Copilot. Both with Microsoft as the main player.

The other option on the market when choosing a data model is Google with its Gemini, which still needs a long way to go to live up to the previous ones.

In Generative EI, we have formats for creating text, images, video, voice, and music. The most relevant thing is that the content generated is not subject to copyright (at the moment). This generates an intense debate because the models have been trained with content that did have copyright, without the permission of its creators.

For Managers

Among the relevant aspects when making decisions about the use of AI in our organization, I would take into account the following:

The use of AI by a few can further increase the digital divide.

The organization must unify usage criteria to prevent each advanced user from using solutions outside the control of the organization.

Data policies must be established that include what can be used on these tools.

Prompts transfer know-how to models. It is necessary to use paid tools with terms and conditions that allow working on private models in which neither the data nor the prompts are shared to train other models.

Users must be trained to reduce fear and avoid excesses.

We must convey to our teams that AI provides tools that will make good professionals better, not replace them.

AI alone does not provide value; The value is provided by professionals who know how to take advantage of their potential.

Interaction with AI is a Human-AI-Human trinomial. The first step is knowing how to write the appropriate prompt. The final step is knowing how to take advantage of the content returned by the tool.

The speed provided by AI accentuates the need for immediacy that currently affects society.

Also in the case of the use of AI in professional environments, the future is not what is going to happen, but what we want to happen.

Also Read: Top 9 Things That Artificial Intelligence Can’t Do

Leave a Reply

Your email address will not be published. Required fields are marked *