Between July and September, the wildfires in the Amazon and Pantanal became the focal point of Brazilian media due to their extensive scale. According to data from the National Oceanic and Atmospheric Administration (NOAA), the smoke generated by these fires spread over a vast area of Brazilian territory, forming an air corridor that carried the smoke from the northern to the southern regions. Mainstream and social media, especially in Brazil, extensively covered the impacts of the fires, highlighting the harmful effects on human respiratory health and the loss of non-human lives in the affected areas. Images of burning forests, orange-hued skies, and wildlife being rescued flooded the media. This scenario reinforced the urgency of real and concrete measures to address the climate crisis. The image captured by the NOAA-20 satellite reveals the propagation of the smoke. The presence of smoke in areas far from the hotspots underscores the severity of the situation. The fires occurred due to a severe drought, which was most noticeable in the Madeira River, and spread through the states of Rondônia and Amazonas, further aggravating the challenges faced by both the ecosystem and its inhabitants.

We propose a collaboration with the artificial intelligence databases that feed the algorithmic image-generation systems. By defining a prompt for tracing the smoke path, we establish a predefined narrative. When comparing an AI-generated image to one captured by a smartphone camera, the complexity of image-generating systems becomes clear. Below is the comparison of two images: First, an AI generated image created with the prompt previously mentioned; The second is an image taken by Fernanda Oliveira on September 19, 2024, in Manaus-AM, using a Samsung Ultra 2023 smartphone camera and digitally edited for clarity.