While the West debates how to limit the use of artificial intelligence on university campuses, on the other side of the world China is organising to overcome the obstacles imposed by US sanctions on the advanced chip front. In between, Valve’s founder announces the first ‘Starfish’ brain-computer interface , ready to subvert human-computer interaction as we know it. Three seemingly unconnected stories that actually tell the same scenario: AI is no longer a matter of software, but of geopolitics, education and biotechnology.
Let’s start with Tencent and Baidu, two Chinese technology giants who, as reported by CNBC, have unveiled their strategies to cope with the US blockade of AI chip exports. Instead of slowing down, they are reconfiguring their language models to operate in less powerful hardware environments. Baidu has even upgraded its Ernie model with a 4.0 version that takes advantage of optimised computing efficiency, proving that innovation can come from restrictions. Tencent is not to be outdone, adopting a modular logic for the training of its LLMs, so that it can scale nimbly even on non-top-of-the-line infrastructures. The move is clear: if they cannot compete on chips, they will do so on efficiency and architecture. A new balance between software and silicon is emerging, and Beijing seems more ready than the West would like to admit.
On the other side of the Pacific, however,artificial intelligence is shaking up the American education system. Axios reports how a growing number of university professors are facing a real emergency related to the massive use of ChatGPT to copy, write papers and even take exams. Some courses have already reverted to pen and paper, while others have dismantled traditional assessments to rely on oral interviews. But what is most worrying is not so much the cheating as the epistemological fracture that is opening up: the one who controls the information is now no longer the teacher, but the algorithm. And it is here that the parallels with China become disturbing. While Western students are using AI to ‘bypass’ study, the East is using it to redefine national cognitive production.
Then along comes Gabe Newell. The founder of Valve, as reported by The Verge, announced the first BCI (Brain-Computer Interface) brain chip, called Starfish, already implanted in a volunteer. This is not a futuristic product à la Neuralink, but a technology that is already functioning and potentially marketable. The dream of controlling video games with thought is only the first step. What Newell is really proposing is the beginning of the post-AI era: when the boundary between mind and machine will no longer be virtual, but biological. The real challenge will then be who will control these devices, and how they will be integrated into the social fabric.
The shift is clear: from the control of computing power, to the manipulation of information, to human-machine hybridisation: artificial intelligence is no longer a mere tool, but an environment, an invisible infrastructure that crosses borders, brains and educational systems.
The real battleground will no longer be between OpenAI and Google or between Nvidia and Huawei. It will be between those who will manage this cognitive transition and those who will fall victim to it. One does not need to look too far ahead to understand where we are going. Just look at the cracks that are opening up today.
What if the next step is not a software update… but an update of our own brains?
