Artificial Intelligence amazes us more and more with its capabilities. Its “brain power” is increasing (good), its consumption is increasing (bad), but its risks are also increasing (very bad). This is not a fairy tale: without a “moral code”, AI could really evolve to the point of endangering human values and well-being. This is why it is time to change the paradigm: from intelligence toartificial integrity.
Only by integrating ethical principles into machine learning systems can we shape a future in which artificial intelligence prioritizes our safety, health and fairness, preventing those who operate it (because it is still a tool) from using it for profit-related purposes, such as news reporting. he's already showing us. It is not an easy path: the contribution of ethics experts will be needed, sociologists, philosophy. But it is a challenge that forward-thinking leaders in every industry must address now. Artificial Integrity is the key to a tomorrow where humans and machines collaborate in new and more ethical ways.
Towards Artificial Integrity: Hamilton Mann's Vision
One of the most brilliant thinkers on these issues is Hamilton Mann, with whom I had the pleasure of exchanging ideas on a couple of occasions. Mann is a pioneer of ethical AI: the creator of the very concept of Artificial Integrity, which delves into in his book "Artificial Integrity: The Paths to Leading AI Toward a Human-Centered Future".
His vision is clear: to build a future where AI is aligned with human values, we must move from intelligence to integrity with serious, practical, concrete steps. This means annotate training data with codes of ethics, supervise machine learning processes e integrate human value models into systems. Only in this way will we be able to have AI capable of making fair, inclusive and transparent decisions.
A multidisciplinary and global challenge
I repeat: Artificial Integrity cannot be the sole responsibility of developers. It requires interdisciplinary collaboration: in addition to the aforementioned sociologists, philosophy and ethicists, legislators and different user groups will also need to be involved from the beginning. And since society's relationship with ethics is dynamic (sometimes elastic, to be precise) mechanisms will be needed for AI to adapt to evolving ethical standards over time. Above all, responsible leadership to coordinate efforts.
An emerging field is thealgorithmic, which studies the ethics of algorithms: teams of humanists and computer scientists work to define moral principles to translate into code1. With the explosion of generative AI, these reflections are more urgent than ever. We cannot allow such powerful technologies to operate in an ethical and legal vacuum.
Four Operating Modes for Ethical AI
Mann identifies four ways in which humans and AI collaborate based on integrity:
- Marginal: AI identifies areas where both its and human input are no longer needed, allowing people to be reallocated to higher-value activities.
- AI-First: Even when AI plays a leading role, it maintains ethical standards of fairness and inclusion. Not only that, it can explain how it arrives at certain results and avoids bias, for example in the medical field.
- Human First: AI supports human decisions without replacing people's ethical judgment, as in judicial processes.
- Fusion: the perfect synergy between AI and human intuition, for example in self-driving cars where AI handles the driving but human oversight intervenes in ethical dilemmas.
Artificial Integrity: The Questions Leaders Need to Ask Themselves
AI will continue to evolve, opening up extraordinary opportunities but also profound ethical and social challenges. Leaders in every industry must ask themselves some crucial questions:
- How to avoid over-reliance on AI where human judgment is essential?
- How to manage the potential increase in employment and income inequalities?
- How to prevent unethical uses of AI and large-scale privacy violations?
- How to regulate AI development in a responsible and sustainable way?
Towards a Democratic and Ethical AI Future
I am cautiously optimistic that widespread access to AI tools will continue to democratize these technologies, making them available to small businesses, innovators, and individuals who can use them for good. The more people have the opportunity to understand and apply AI, the greater ethical awareness and pressure for better governance will grow.
Even as we race toward a future shaped by AI, we have the opportunity and responsibility to put integrity at the center of it. Only then can we reap the extraordinary benefits of AI while minimizing its risks and building a more ethical and inclusive society. There is still a long way to go, but with touchpoints like those outlined by Hamilton Mann, we have a map to orient ourselves.
The future of AI is, must be, integrity.
- If you want to delve deeper into the topic of algorithmics, also in relation to the concept of Artificial Integrity, I invite you to follow Paolo Benanti's work, among the major experts (and certainly the most passionate) on the subject. ↩︎
