Artificial Intelligence is used to make predictions. If we want to unleash its full power we must rely on this force.

When we consider the power of electricity in specific applications, we aim to improve the strength or speed with which tasks are performed. For example, electric printers allowed us to print newspaper runs more quickly compared to mechanical printers. Similarly, by replacing a manual sprayer with an electric sprayer, we avoid carrying a heavy weight on our backs and using arm strength to spread fertilizer.

Nevertheless, the great advantage of electricity does not lie in point applications, but in observing the advantages it offers as a complete system. In the book Power and Prediction, by A. Agrawal, J. Gans and A. Goldfarb, an even more important idea is put forward: electricity has spatially decentralized the use of energy, which means that it is possible to apply more force in a process no matter where the person applying it is located.

Thousands of MSMEs in Colombia remain afloat in the face of strong economic tensions

Before electricity, industrial processes were centered around a steam boiler, and the farther away you were from this engine, the less power you could apply. With electricity, however, it doesn’t matter whether you’re on the third or tenth floor, in the countryside, or inside a car, because in all cases you can power your devices with the same power and energy. You have simply unleashed your ability to perform tasks anywhere.

Similarly, oil represents a revolution when we look at the system as a whole. The way we harness the chemical energy of its derivatives makes things more durable, which in turn frees us from having to repeat the same tasks over and over again.

For example, gasoline cars overtook electric vehicles more than 100 years ago, due to the charging time and the vehicle’s autonomy on each charge. It only takes a few minutes each week to fill up a gasoline car, compared to the hours it takes to charge an electric car.

Likewise, plastic bags emerged as a solution to the felling of trees, since the production of a paper bag 60 years ago required several decades (considering the growth time of a tree) and had only one use, unlike their oil-based counterparts.

Now, the question arises: What is the advantage of Artificial Intelligence systems? If we analyze the specific applications of artificial intelligence, we can note that Google provides us with information more quickly, or that ChatGPT writes texts more quickly.

We could think that implementing artificial intelligence systems in our company is a way to speed things up, but this would be a misconception. In fact, the task of streamlining processes has been achievable for decades with automation systems, algorithms, and robotics.

The real advantage of Artificial Intelligence systems lies in their predictive capacity, in the reduction of uncertainty about what will happen in the future. When you consider applying Artificial Intelligence in your company, you should ask yourself: What aspects of the future do I want to know?

If you have an agricultural business, you might ask yourself: How many tons of production will I have next month? If you have a convenience store, you could ask yourself: what will be the sales of notebooks during the next school season? If you are a financial institution, you might ask yourself: how many and which clients will default on their loans in the next 3 months? And so on. It is important to understand what forecasts you need to know and rely on the information provided by Artificial Intelligence to make decisions.

However, switching from one system to another represents a considerable effort, as companies are currently built on standard procedures, checklists, and rules that employees must strictly follow.

An AI-based, prediction-oriented system has a different logic and does not operate under rules where the answer is simply “yes” or “no”. This can lead to friction with existing corporate structures.

Additionally, there are risks and fears in making a change from rule-based to forecast and probability-based systems, some of which I will discuss in my next posting!

Note: Proofreading of this article was done by ChatGPT.

From electricity to Artificial Intelligence