Artificial Intelligence provokes highly contrasting feelings in those without a deep knowledge of it: from strong attraction to profound repulsion, some are seduced while others completely shun it.
AI has an enormous memory and manages very well calculations or the recognition of images; it adheres to data; it’s rational and objective. When, however, you ask it to be creative, that, in our opinion, is when shows its limits if, by creativity we mean unpredictability. Here we need human ingenuity, which is subjective, guided by motivations and emotions but which, with empathy, is able to find solutions involving different concepts that have never been combined before.
For this reason, we prefer to talk about augmented intelligence, a combination of AI and human intelligence, but we use the term, “Artificial Intelligence” because it’s widely used throughout the world.
Is AI really intelligent?
We’re not really convinced that we’re talking about intelligence. It simply carries out what a human tells it to do. The human is intelligent because knowing how to use AI is not easy.
Artificial Intelligence is today an inflated term, it’s often used in inappropriate contexts with negative connotations, sometimes justifiably when used on drone killers.
For this reason, we feel it necessary to share with you how we at Hopenly use it.
Like all technologies it’s not in itself right or wrong, dangerous or life-saving; it all depends on how you decide to use it. We believe it’s necessary and fundamental to declare our commitment to develop ethical and transparent technologies.
For us at Hopenly, for its ethical development, Artificial Intelligence must be based on the following values:
AI is a great resource and an opportunity for many sectors. For this reason, it’s not enough to educate Data Scientists and Data Engineers to adopt an ethical use; it’s necessary to train and question things continuously.
We believe that algorithms should be developed with the objective of strengthening the ability of people to make decisions based on data, while leaving the opportunity to intervene in every decision arrived at with Artificial Intelligence.
Our AI algorithms are developed in such a way that the client can understand their functioning, what data are used, and how.
Guaranteeing, moreover the traceability of results allows people to examine the models and understand how they work, in order to verify the output and understand how any errors can be corrected.
The algorithms that we develop for our clients cannot substitute the decision-making of human beings. We always point out their capacities but also their limits, precisely because our objective is to create an AUGMENTED INTELLIGENCE.
Diversity and Inclusion
Cognitive biases constitute a danger for Artificial Intelligence and could have negative effects such as marginalization, an increase in prejudices and discrimination. Having a diverse team by age, gender, ethnicity, culture and background allows us to avoid them and create more reliable algorithms, the variety of points of view being a protective factor against biases.
When we talk of responsibility in the development of AI algorithms, we mean both our responsibility in the development of the algorithm, and responsibility regarding its consequences.
This concept, considering the ethical guidelines of the European Commission, also extends to accountability; besides being responsible, we also need to be able to justify our choices and those of the algorithm.
In Hopenly, an independent team with respect to the team dedicated to AI development evaluates the risk and validation of the algorithms, verifying that no model contains problems or is used in an inappropriate way.
Social and environmental wellbeing
We consider AI to be a tool that generates wellbeing, especially if used to improve the quality of life. We also work in the medical sphere, the sector that most benefits from the new technologies. This, for example, is a project that we have worked on with the cooperation of the Baggiovara Hospital (MO): Seasonal variation of semen parameters correlates with environmental temperature and air pollution: A big data analysis over 6 years.
We use sustainable technologies to reduce environmental impact and create value for future generations.
Visit this page to find out more about our commitment to sustainability.
Data reflect the culture of the context from which they have been extrapolated. Many examples of AI also assimilate people’s prejudices together with the information extracted from the data.
We strive to free data from historically sedimented prejudices and from cultural stereotypes, which people are often not aware of or don’t believe can be found in their data.
In this way we can construct impartial AI systems for taking decisions.
Security and privacy
Data are the fuel of AI, and we work with any type of datum, including sensitive data when developing algorithms in the medical, insurance and banking fields. In 2018 we followed the procedure for becoming G.D.P.R. Compliant, guaranteeing our clients maximum security for their data which we need to analyse. Access to client data is reserved to people authorised for processing, identified when entering into the contract.
We ensure the maintenance of the quality and integrity of data, thanks to our platform, which allows us to monitor and maintain the algorithms developed and the data bases which we work on.
We believe in the combination of Artificial Intelligence and human intelligence, in which the limits of one are compensated by the other in a union that creates value.
For this reason, we undertake to use technology according to the ethical guidelines of the European Commission, to create Artificial Intelligence which is always supervised by human beings in every phase and which follows the regulations regarding privacy and the protection of data.