The accelerated growth of the volume of data available today is driving the emergence of a variety of new applications aimed at solving various problems affecting our society. Among these sources of information, satellite images occupy a prominent place because they allow the identification and analysis of various environmental phenomena such as forest fires, floods, droughts, water leaks or land use changes, among many others.
When the use of artificial intelligence (AI) is integrated into this type of data, particularly through machine learning – a branch of AI based on learning patterns from examples – its application potential is virtually limitless.
Given this scenario, a key question arises: Is the use of these technologies reliable for solving real-world problems? In most cases the answer is yes. However, AI-based models have an important limitation, particularly in the area of public policy and social problems: the lack of explainability. Although these systems can operate with a certain margin of error, their use can become problematic and create distrust if it is not possible to understand why certain decisions are made.
In this context, the value of AI depends largely on understanding the problem you want to address. It is not always convenient to apply intelligent models in every situation, as their usefulness depends on the specific circumstances, the availability of reliable data, and the transparency requirements that the solution requires. For this reason, the debate about the indiscriminate use of AI remains open.
Although satellite imagery is usually relatively easy to obtain and in many cases available for free, in Latin America there is a heavy reliance on private companies and organizations from developed countries to access this data. This situation limits the region’s ability to develop innovative applications and creates significant vulnerability, as there is no guarantee that access to such inputs will be maintained in the future.
Unless Latin America manages to develop its own satellite technology that can produce high-quality and accessible data for the region, any solution that combines AI algorithms with large databases will continue to be dependent on third parties. However, there are already numerous applications that demonstrate the value of these tools, such as estimating poverty rates from nighttime images, detecting water leaks, or even identifying secret graves with profound social impacts.
Despite the limitations and risks, these technologies represent an opportunity for growth and development. When applied in a responsible, ethical and conscious manner, complex problems can be solved more quickly and efficiently, positively impacting the daily life of society.
Copyright Latinoamerica21.com
About The Author
