Banco Sabadell and FADE (Federación Asturiana de Empresarios) organised a conference to analyse the challenges, perspectives and also the risks that companies face with the arrival of Artificial Intelligence.
Alfredo Fernández Santos (North-west Sales Director of Banco Sabadell) and Leticia Bilbao Cuesta (Head of the Corporate Development Department of FADE) were present at the one-day seminar and the guest speakers were Irene Díaz Rodríguez (Chair of Computing and Artificial intelligence) and Luis Echavarri Lasa (Head of the Analytics and Artificial intelligence Unit of Banco Sabadell).
A technology with unstoppable growth
During the day, some data were presented that highlight the great speed at which the use of Artificial Intelligence is evolving: “It took Twitter 65 months to reach 100 million users, while ChatGPT managed to reach that same figure in just two months”. Experts predict that the impact worldwide that Artificial Intelligence will have on GDP, from today until 2030, will be of magnitudes never seen before. Currently, it is estimated that 20% of China’s GDP is already directly related to AI. An equally exponential growth has been detected in the business world. Thus, if in 2021 only 7% of companies used some type of Artificial Intelligence, today, and in the absence of official statistics, it is easy to assume that the vast majority of large corporations already have AI initiatives on the table to address this in the short-medium term.
Vueling’s chatbot, one of the best examples of AI in the corporate world
In her talk, Irene Díaz recalled that this technology is not new. It has been around for many years now and, little by little, it has been incorporated into many of the applications with which we interact in our daily lives, including, for example, household vacuum cleaners. The “great revolution” has been marked by the arrival of Generative AI whose main protagonist is the aforementioned ChatGPT.
As she recalled, in some cities in the United States, there are already companies that use “riders” or fully robotised delivery drivers or driver-less taxis. These are, however, phenomena that are still in an embryonic state, given that these solutions must be developed in environments of advanced interconnectivity such as the well-known Smart city.
On other levels, this new technology has led to the emergence of multiple tools that facilitate the work of creators in fields as diverse as music, painting, design, marketing and writing. There are already companies that have managed to implement, with remarkable success, AI systems in their structures, significantly improving the level of service they provide to their customers. As an example, she cited Vueling’s chatbot which, she said, possibly provides its users with a much better level of service than much of its current operators could provide.
“There are already Artificial Intelligence systems for any type of business management”
At the moment, he said, companies already have within their reach the possibility of designing AI systems that allow automating, or improving, many of the business areas; from the creation of new products, to the design of personalised offers or implementation of quality systems.
However, the implementation of this type of solution is not an easy task nor is it without risks. These systems do not work on their own. They are fed by immense databases that have previously had to be fed and must contain good quality reliable data. “It can be much more complicated to get the reliable database we need than it can be to design our AI system,” she said. “If the data are not good, the system will fail”.
Another common mistake, she went on to say, is to think that an AI system will do everything for us. This is false. These types of tools have to be considered as “helpers” or “auxiliaries”, hence the well-known name “co-pilots”, that facilitate people’s work, never as substitutes for them.
Pending Issues in the field of AI
In this same vein, she mentioned some of the main issues yet to be resolved in this field. Referring more specifically to the world of medicine, where AI already has multiple applications, she pointed to the fact that doctors already have tools that help them make decisions, but they demand knowing what routines and sources of information systems have been used to provide them with the advice they receive. “In other words, it is useful for any doctor, before making a diagnosis, to have a document that provides them with a summary or synthesis of hundreds or thousands of similar diagnoses made before by other doctors. What we need now is to report also on how this information has been prepared”.
Another risk to consider before deciding to develop an AI system in a company is the absence of legislation that regulates it. Progress is being made very quickly in this regard, especially in the European Union, but technology is currently moving faster than legislators, and this is something that must be borne in mind.
What should a company consider before launching its own AI system?
For Díaz, any company, before deciding to implement one of these systems, must ask itself the following questions:
- Do I have the necessary resources?
- Do I really need it? How will it benefit me?
- What risks may I be taking?
- Above all…do I have the necessary data? Without quality data, there is no Artificial intelligence.
She also referred to a factor that is often overlooked: the data that companies have is usually historical data. Therefore, they are biased and were compiled at a time when society was not the same as today’s. Thus, for example, it is very possible that data collected thirty or forty years ago will give the male gender a weighting that is much higher than it should be today.
AI, a highly polluting technology
Finally, she recalled another topic that is not talked about much: AI is a highly polluting and environmentally harmful technology. Its consumption of water and electricity is very high. Technologies such as ChatGPT, both its development and its use, pollute more than a fleet of cars that are more than 15 years old. This is another key aspect to consider as technology advances.
You can read the main conclusions of Luis Echavarri’s speech on AI in the financial sector and its applications at Banco Sabadell here.
