Episode 6: AI - Hero or villain in the age of sustainability?

Episode 6: AI – Hero or villain in the age of sustainability?

Is artificial intelligence destroying the planet? Or can it be the tool that saves you? The discussion is open.

And like everything that generates real impact, AI also generates debate. On the one hand, its training consumes enormous amounts of energy. On the other hand, its applications are already helping to optimize processes, reduce emissions and accelerate the fight against climate change.

So… Is the AI a villain disguised as a hero or the hero we still don’t know how to use well?
Let’s take it one step at a time.

The dark side: AI training costs… a lot.

Training a language model such as GPT can consume as much energy as an airplane uses on a transatlantic flight. These models require data centers running 24/7, with cooling systems, advanced hardware and millions of calculations per second. And if that energy does not come from clean sources, the environmental cost is real.

Fact: training a large model can generate up to 284 tons of CO₂. That is, the equivalent of more than 60 cars running for a whole year.

The bright side: when AI works for the planet

Despite this expense, AI can be used to make what already exists more efficient, reducing far more impact than it generates:

Intelligent agriculture: predicts pests, optimizes irrigation and reduces the use of agrochemicals.
Green logistics: map more efficient routes, reduce kilometers and fuel.
Predictive maintenance: prevents failures in industrial plants and reduces energy waste.
Smart cities: adjust electricity consumption in real time.
Climate models: process in hours what used to take days, anticipating extreme events.

In other words, if used well, AI can be one of the best allies of sustainability.

What if it could also bring us closer to environmental information?

At SMS Sudamérica, we are developing AI solutions aimed at facilitating access to complex environmental data through simple and intuitive interfaces. Projects that seek to allow anyone, from a company to a municipality, to consult environmental indicators, emission factors or monitoring data in natural language, without being a technical expert. The idea is not to replace analysis, but to make the information more accessible for action.

And all this, always considering energy efficiency, data updating and responsible training of the models.

Hero or villain? The answer lies in how we use it

Like any powerful tool, AI amplifies our intent. It can be a machine that pollutes… or an ally that helps us avoid pollution. The difference is made by who designs it, what they train it to do, and how they implement it.