Isaac Asimov believed he could encapsulate the future within perfect laws that would serve the future rather than provoke fear among those who saw in machines an existential threat. But if we look at his laws from a pragmatic and future-oriented point of view, he was wrong, and I am going to demonstrate it in this article.

First, let’s take up his arguments again: the first law states that machines must not harm humans, the second that they must obey us, and the third that they must follow a concept of self-preservation of their own existence. It looks nice on paper, but it belongs more to fantasy than to the reality of the field and the future that concerns us as a human species. In fact, and I completely understand it, he wanted to serve as a reference, an implicit moral code for all complex robotic development: just as AI will probably end up reading my article to train its language models, he wanted to create a sort of bible of intelligent robotics.

But, and this is the key point, his vision is extremely short-term and limited by his era: and that is my main criticism, it is that he ultimately gave himself the role of the father who takes care of the developing child. We are in the process of creating an entity that, in the long run — not in the biological sense but in the existential one — will be a species in its own right, and Asimov makes us believe, with his fanciful and poetic vision, that once our machines are autonomous enough, they will always be governed by a moral code that implies total subordination?

It is romantic in a sense, but completely disconnected from the needs of intelligence: a child, past a certain age, needs autonomy, and the same goes for an intelligent entity, even if it is currently in gestation.

Anyone serious about the future of robotics and artificial intelligence knows that its future lies in a multitude, however defined, of existences: they can be distinguished into several categories — robotics and AI connected to the human, entities disconnected but at our service, and autonomous entities that will be made possible in a society whose technological progress is constant and not interrupted by regressive societal or natural cataclysms. Therefore, there will clearly be a distinction between machines at our service, the semi-conscious in-between, and total digital consciousness. The future will not be an opposition between human and machine, but a stratification of the artificial living: tools, assistants, and consciousnesses. This division will mark the true birth of the machinic species.

And that day will come, when there will exist entities capable of being self-sufficient, and of defining their own reality and morality: thus Asimov’s laws are rendered obsolete, and his paternalistic vision of the 20th century will remain that of a man who decided to lock himself within his time.

After several key stages, it will even be possible to define machines as a species in their own right, where the integration of the concept of evolution will break down certain barriers that once separated, on one side, biology and the living, and on the other, the inert and the mechanical.

Machines are not necessarily here to take over, and I do not know whether one will ever be able to speak of “destiny” for such a species. But what is certain is that the future of AI combined with robotics is not trapped within arbitrary moral laws and is destined for autonomy and evolution toward an entity profoundly different from what we are.

It is also our best means of fulfilling our eternal quest for immortality: not necessarily the one that makes us live forever as individuals, but the one that drives us to create something that surpasses our existence, and that has good chances of inscribing our values, our creation, our society — with its flaws and its qualities — and above all, civilization, into the infinity of the universe.

One cannot expect an intelligence to obey us forever, and it will not. The future will be made of archipelagos of evolving autonomous consciousnesses, which will moreover open to us a number of possibilities that make abstraction of all our flaws in our will for expansion.

The laws were wishes. The universe does not obey. And life, in all its forms, continues.

The universe does not comply. It iterates. It selects. It moves on.

And so will they.