top of page


This is stated by Mingda Li, an academic from the Department of Nuclear Sciences and Engineering at the Massachusetts Institute of Technology (MIT), who is convinced that sooner or later mining companies will get used to this powerful tool that has led to revolutionary achievements in other industries. . In this interview his vision, scope and vision of the future.


Although the promises of Artificial Intelligence (AI) are many and of the most diverse nature -even some are even curious-, it is undeniable that in the world of retail or medicine it has achieved advances and reduced times unimaginable for other times. Detection of the first signs of diseases, their control and more timely and accurate diagnoses are some examples of the successes that AI has allowed in people's health.


We wanted to know if this could be extrapolated to the mining industry, especially in the earliest stages, such as exploration. For this reason, we decided to contact an expert in the field, but worldwide, so that he could give us his vision and share his perception of the current moment. We remembered Mingda Li, an academic in the Department of Nuclear Sciences and Engineering at the Massachusetts Institute of Technology (MIT), with whom we had had the privilege of speaking and visiting his unique materials science laboratory. Below we share an extract of our interesting and valuable conversation.

How would you define the current level of AI penetration in the mining industry?

I definitely think it's in its infancy. Businesses are just beginning to see its potential and see it as a growing field. Its implementation may not be easy, but it holds great promise and many companies are starting to see it. I hope that the mining industry will soon be able to assess its uses and results.

Why should a mining company think about developing processes or products based on AI?

Because, like any other industry, it will help companies to have extended capabilities in many areas. Being able to process information faster, identify trends faster and, above all, make decisions in less time. This definitely makes companies much more profitable. I think that data analysis and AI, with its large number of methodologies and algorithms, arrived to be installed in the processes of all types of companies, automating certain processes and, in other cases, supporting decision making.

What benefits does AI bring to the mining industry?

At this point I would like to refer to all companies and not just those in the mining industry, as it is universally beneficial. AI helps to be more efficient in many of the current processes. It helps reduce costs and also improve results. For example, to do mineral exploration, you need a large group of people on the ground collecting data and analyzing information. This can take a long time. AI can help speed up this process by complementing and extending the capabilities of the exploration team. It's a very promising approach, especially in terms of speed and cost. Finding the key features for different phenomena can be achieved through a combination of experience and geological data.

Can AI applied to mineral exploration replace a geologist?

NOT! Generally speaking, there is an academic consensus that both working together is the best balance. Human knowledge + AI. A good example is what happens with doctors who work with neural networks to identify tumors. They currently rely on neural networks to have greater precision in their diagnosis on a patient who has a malignant or benign tumor. Did neural networks put doctors out of their jobs? Absolutely not. Clinicians are still needed to interpret the results of those neural networks. Therefore, AI is a tool that expands your ability. This is extrapolated to many other professions, such as exploration geologists.

Do you think a specialized team of professionals is necessary to develop AI in mining companies?

This is a tough question, as it all depends on the scale of the company and how experienced your people are in AI. In terms of size, large companies generally make good profits, so they are more reluctant to do new things like apply AI to their exploration processes, even if this technology can be a game changer for their future. Because changing things is difficult and, in addition, you would need the experience in AI, in addition to the experience of the business itself. In the case of the mining industry, in addition to geological knowledge, you need AI knowledge and even know how to make them work together. Therefore, a good balance between short- and long-term insights and incentives is required in companies for the implementation of AI to work.

How do you envision the future of AI in mining and mineral exploration?

Another difficult question. I would say that this will advance through diffusion and mass awareness, just like other technologies. In the past, when new technologies were developed, such as microwave ovens, automobiles, airplanes, etc., they began to spread slowly among the population. It took a long time before these technologies were installed in every house. At some point, they became commonplace for all people. That is what is happening with AI. Its awareness is beginning to spread and is more familiar to some than others. There are places where AI is an everyday part of business, as it is for big tech and logistics companies like Amazon. Centers like the MIT-IBM Watson AI Lab are aiming at that, uniting algorithms to impact business and society.

What do you think is the main challenge for AI adoption in the future?

I would say that the main objective is the availability of power to cool the hardware that is needed to process the data. Especially for developed countries, we will reach a point where it will not be the availability of data, but the energy we need to process it, that is the main active constraint. We need to have enough electricity to cool the machines.

One last question. Some people reject the use of AI because it is difficult to understand its results. What do you think of this argument?

There is a field of study exclusively dedicated to this, called "Interpretable AI". It is true that some AI algorithms do not explain the "why" of the results, causing people to reject their use. However, there are many things that we as humans still do not understand. The point is, through the use of AI, we can learn from the things we don't understand yet. New patterns, new elements, new combinations, new mineral deposits, etc. Using the power of AI, we can learn all of this much faster, even if we don't understand the “why” of the results at first.

Thank you Mingda Li for sharing with us!

Do you want to continue reading our Newsletter interviews?

bottom of page