Artificial Intelligence is no longer a futuristic technology that was only used in R&D, scientific organizations or military applications. Today we use AI on a daily basis and it is fully integrated into our lives. Our mobile phones use AI to take better photos, we use search engines that predict what we are looking for based on our previous searches or we use recommendation algorithms like those of Netflix, Amazon or Spotify.
In spite of its presence in our lives, is Artificial Intelligence available to everyone?
An early 2020 study by the European Commission shows that while awareness of AI is high among European companies (78%) only four in ten (42%) have implemented AI-based solutions and 40% state that they do not make use of it and do not intend to do so in the next two years (2022).
The main barriers for these companies to adopting AI-based solutions are the difficulty in recruiting new staff with the right skills (57%), the cost of the technologies (52%) and the cost of adapting their own operational processes (49%). This leads most companies to purchase off-the-shelf solutions (59%) or hire external providers to develop them (38%).
Therefore, while Artificial Intelligence is within everyone’s reach and we use it on a daily basis, companies are not adopting it as widely, either due to a lack of suitable personnel or due to internal operational frictions that make the implementation of AI-based solutions impossible.
Understanding what kind of problems it should be used for, how to use it, and how to understand both how it works and what the results will be is critical to the success of these applications.
Is it therefore possible to train in Artificial Intelligence in order to facilitate its future use?
AI is built on three main pillars, mathematics, programming and computational skills. While the first of these requires more extensive and formal training, the great advances that have been made in the field of programming mean that the development of an AI solution can be carried out in a simple and efficient manner.
With a basic knowledge of Python, anyone can develop an image classifier in ten lines of code, but understanding what is happening for our classifier to work correctly requires previous knowledge and many hours of specific training and practice. This is one of the most significant barriers to adopting Artificial Intelligence.
In addition, the expansion of the internet, and its use as a tool for sharing knowledge, makes it easier and easier to learn about these new technologies and their development. If we look at the increase in searches for programming courses (red) or machine learning courses (blue), one of the most widespread AI disciplines, we can see how they have increased over the last ten years.
Meanwhile, computing power has been growing and any mobile phone has more computing power than NASA had during the Apollo 11 mission calculations. This growth has enabled AI to expand and AI-based solutions to be adopted.
Therefore, there are ample resources for training in Artificial Intelligence techniques, but without adequate mathematical training, there will still be barriers to AI adoption due to the difficulty of interpreting the techniques and their results.
Finally, is it an advantage for organizations to have staff with AI skills?
It is important, but not essential. Having trained personnel will allow mainly the 40% of organizations that have not yet developed AI-based solutions to make better decisions when it comes to making the leap to using them or the 38% of organizations that turn to external providers to distinguish which ones are the right ones for their needs.
Meanwhile, at Numens we are experts in Artificial Intelligence, and we can help you identify the adoption and implementation opportunities that these technologies could have in your company, select the right technologies and suppliers, and understand both the performance and the results of applying these techniques in your business.