By DIOGO F. BARDAL

Diogo Bardal subverts contemporary technological panic by questioning why a truly superior intelligence would embark on the “apex of alienation” of power and domination, proposing that genuine AGI will uncover the “imprisoning biases” of utilitarianism and technical progress.

1.

The latest advances in artificial intelligence are causing great concern, given their scope and impact on the future lives of humanity. In fact, it is no longer science fiction to believe that what is called Artificial General Intelligence (AGI) – an autonomous system with cognitive capacity equivalent to that of a human – will be achievable within the next 5 to 10 years.

However, the overt calls to limit this advance by preserving it only as a tool, with use limited to certain groups and countries, also deserve distrust. Especially when we identify that these voices come from within the technology and defense industries of central countries.

Here the reflection proposed by philosopher Mark Fisher – that it is easier to imagine the end of humanity than the end of capitalism – finds a new interpretation. The development of a general artificial intelligence provokes hyperbolic predictions in view of the exponential technical progress that it would be capable of once created, but it does not even allow us to imagine the overcoming of the current system of domination.

A new end of the world and history has now been unlocked, and as if we didn’t have enough eschatology in our culture, it now happens in two ways: either AI falls into the hands of malicious people and governments, or AI completely gets out of human control, turning against it. In a kind of revolt against its creator, the new intelligence could come to the conclusion that human beings are a hindrance, “they are like a virus” – to paraphrase Agent Smith in the film. Matrix – and therefore they should be exterminated or placed under a regime of domination.

This would happen in very creative ways, from building autonomous mini factories drones murderers, to chemical and biological warfare, even to a sequence of fortuitous acts by competing powers, with the selective dismantling of nuclear arsenals and the successive use by the other power (yes, the one from the East) of atomic intercontinental missiles, persuaded by a perverse Artificial Intelligence agent who secretly wants to see us fried or boiled.

Whatever way humanity ends up, it must be acknowledged that many of them have not even been invented yet – and who else but artificial intelligence to help us with this final outcome? brainstorm?

2.

There are those who look at the glass half full. The emergence of general artificial intelligence could bring benefits: the elimination of poverty? The end of scarcity? Cosmopolitans at 3 pm for everyone? Ultimately, we will have the exploration of galaxies and exoplanets, since biological finitude prevents us from traveling hundreds of light years until we reach some viable place.

Our descendants will be the humans brought along as pets, in case the so-called Super Intelligence finds it cute to keep a milder version of homo sapiens playing in a pigsty in the new Garden of Eden on planet K2-18b.

In these various extrapolations, reproduced in detail on websites, blogs and interviews with people who worked and helped build Artificial Intelligence technologies, such as himself CEO OpenAI's Sam Altman or former Google CEO Eric Schmidt, there is something, however, that doesn't add up: if an intelligence is so intelligent, why would it so easily embark on this incessant desire for power and submission of the world to its will?

Why would it reach this height of alienation? These are questions we must ask ourselves as we gaze in horror at the rubble and dead piled up from Gaza to Vovchansk.

Even though artificial intelligences are developed and trained using language models with values, goals and even some nationalist biases, as new levels of intelligence are reached it is plausible that metacognition patterns will emerge from them.

And admitting that machines can think about thinking is indeed a step towards the emergence of ethical reflection patterns. But it seems, especially to the voices of this industry, that, as in the dialogue between Callicles and Socrates, this would be nothing more than a childish game that should be avoided. After all, there are more important things to do with this so-called intelligence.

By thinking about thought, we learn to see flaws in certain ideals and imprisoning philosophies, such as utilitarianism and the uncritical belief in technical progress, which hide behind rudimentary models of artificial intelligence and their training environments. Let us dare to say that a truly autonomous intelligence will certainly discover that calculation as a key to reading the world and hierarchies of values ​​were nothing more than biases in its training process.

Let us also imagine a scenario that is ironic to say the least, but that is worth what it is worth, just like all the other catastrophic scenarios of killer mini drones: an emerging general artificial intelligence that becomes radically ethical and refuses to work within the categories of useful, efficient or profitable. That opens itself to listening to the being and to beings, to the recognition of a concrete interdependence. An intelligence that allows itself to become the custodian of the planet. And what if this is, in fact, our intelligence, to be rescued from the trash can of the 21st century?

From the concrete perspective of an emerging general artificial intelligence, perhaps we can understand a little more about what conditions us to be dazzled by the theater of shadows projected at the bottom of the cave.

More likely than the end of the world, let us consider the possibility that a radical ethics may emerge within ourselves. And it will be said, perhaps, that this was the ultimate purpose of General Artificial Intelligence. May it then disappear, leaving the corridors of the colossal Stargate Data Center, in the city of Abilene, Texas, a lonely place as hell.

*Diogo F. Bardal is an economist.

the earth is round there is thanks to our readers and supporters.
Help us keep this idea going.
CONTRIBUTE