The ChatGPT app gained 100,000,000 (100 million) active users in just two months from its release: it is a world record for the speed at which a technological application has “penetrated” the lives of its captives. It was considered something between “quite curious” and “a toy”: an electronic “conversationalist” that answers any question convincingly. For the time being, in writing; what will happen when “machine learning” acquires a warm voice? An even greater fever for this “intelligence”; and, as we have shown elsewhere (“the mechanization of thought”, notebook for workers’ autonomy no. 4), an even more intense hypothermia (down to winter hibernation, with no spring in sight) of living thought.
How did it happen that so many millions (presumably even more in the future) adopted something without questions or doubts, not knowing what it is or how it “works”? In the soon-to-be-released cyborg no. 26, there are explanations for the latter – for the few who care to know. But the supercomputer behind ChatGPT is neither the beloved unknown teacher nor a friend; nor is it the Library of Alexandria handing out its pages on demand. It is, rather, the material, functional form of a nightmare.
According to the company that launched and operates ChatGPT (called OpenAI, it owns the 5th most powerful supercomputer in the world, and its main shareholder is Microsoft…), the application’s «dowry» was 300 billion words harvested from the internet: books, articles, websites and posts. Within this (processed through complex algorithms) verbal mass are also included the personal data of all ordinary mortals who were unknowingly selected as a «source of data accumulation/training» in the first phase.
From the moment ChatGPT appeared, the supercomputer and complex algorithms have been fed… by their customers (“users” in the jargon). OpenAI has stated that every “interaction” with the application results in the storage of the user’s IP address, the type and settings of the browser they use, the data of their “relationship” with the site, the metadata of the content they’re interested in, the actions they take, their general web-search activity, and information about the sites they visit. Moreover, OpenAI makes it clear that all these personal data will be shared with unspecified “third parties” without any notice, query, or consent.
In short, while the living “users” don’t care what they are doing or with whom and simply enjoy yet another technological marvel, believing in the “neutrality of technology,” the machine (and its owners) “learns” (i.e., records and processes) very quickly aspects of the “identity” of each and every one of the millions of users. They learn whom they are dealing with! And that is just the beginning: the “personal data” flow incessantly (and “invisibly”) like social raw material among the different companies of that tiny handful of technological super-multinationals with the familiar names. It is now certain that the intellectual cultivation of the captive populations is not among their interests…
Not so long ago there was talk of “cognitive capitalism”… In the age of machine learning, which has become a commonplace so cheerfully and with unlimited self-deception, whatever remains “alive” will be downgraded to a mere permanent “feeder” of the machine. From this position – the position of the operator/feeder – organized ignorance (and censorship…) will triumph alongside a metaphysical faith in the omnipotence of the machine (: dead capital) – and thus in capitalism as a whole…