This night (being always sleepless, and thinking that Jannik Sinner sleeps ten hours a day), reflecting on the latest video of the Enkk Youtube channel (which is actually called Enrico Mensa, researcher and popularizer with a rare ability to speak of artificial intelligence without chasing the saving or apocalyptic rhetoric, I wrote here), I focused on a matter that seems trivial and in reality he defines the whole game in reality in reality in reality. Course: the world, in the field of artificial intelligence, is divided between “good” and “bad”, and there are only two contenders. A primitive but effective distinction (he says: “I let you decide who the good and who the bad guys are, even if it is clear to me”).
The good, in this case, are the United States and in general Western democracies, because they are structurally bound to slow mechanisms, public debates, procedural guarantees, balance between rights, commissions, moratoriums, ethics, committees. The bad guys, on the other hand, build models. They do not ask for permission or wait for any parliamentary approval, they do not even wonder if it is appropriate to accelerate, they accelerate and that’s it. Enkk says it clearly: three months of advantage is enough. Three months of early access to a self-training model, three months in which a machine learns to improve without supervision (they are already doing it to the Westerners, only that we do not know where they are oriental ones), three months in which we can apply a Darwinian, evolutionary model, for the improvement of the AI, they become an advantage that does not reduce, but amplifies it to dysmissima in more than each day. The distance, like an organism that evolves alone while the other wonders if it is really convenient to evolve.
It is something that, historically, we have already seen in another form. When Reagan understood that the Soviet Union would never surrender for ideological exhaustion, he changed the battlefield: he did not try to win with diplomacy or with morality, rather with economic pressure, the rearmament, the space, putting the Soviet communists in the condition of having to chase something that could no longer afford. So the Soviet Union fell, not because someone killed it, but because he could no longer keep up.
Today Russia still exists, with its old weapons and its financial methods and its expansionist aims that threaten the West and keep Ukraine under the bombs, while on the other front, that of AI, China has everything that was missing from the USSR: a strong economy, an efficient authoritarian structure, a centralized technological vision and above all no remora. Deepseek, for example, as we have seen, built a model comparable to chatgpt with a fraction of time, resources and infrastructure. It doesn’t matter if less elegant (and with several censorships, obviously because controlled by the Chinese government, I bite make some questions about Taiwn), the important thing is that it exists, grows, works, and that they did it. Trump can be accused of anything except slowness and has announced Stargate, a 500 billion dollar plan to equip the United States with an infrastructure to the national teams, with gigantic data centers, autonomous energy production and models capable of supporting American algorithmic supremacy (among other things, one of the reasons for Musk’s anger, since there are Openai, SoftBank, Oracle and Oracle. Mgx, not him: but wasn’t Elon priority to go to Mars?).
Anyway it is: and Europe? Europe is known, looks, regulates, writes guidelines, produces ethical framework, and in theory it is only wrong. Except that the ACT is its monument to irrelevance: a legal apparatus built to control something that has never created. The EU deludes itself to manage artificial intelligence as a dangerous chemical while in the meantime others treat it as a strategic matter. The EU discusses whether to raise military expenses and always comes to us a step afterwards, with the brake pulled, as if the deterrence was an option and the war a phenomenon still negotiable with diplomatic notes. The pacifists (or “pacifinti”, as they are called today) have not understood that armoring means deterrence, and deterrence preventing war (see the case of Ukraine again).
Last week he said it to Guido Crosetto, defense minister and friend, and he replied: “I have been repeating it for months, it is difficult to make it clear”.
In any case, the problem of the vouchers and bad guys in the best racing raised by Enkk remains, and it looks like a cul-de-know: either you put limits, and stay back, or do not put them, and win. The risk is in both cases.