Your belief that "power structures can't be fixed" perfectly illustrates what educator Paulo Freire described as the oppressed having a "diffuse, magical belief in the invulnerability and power of the oppressor". Anthropologist David Graeber noted that modern capitalism has constructed a vast bureaucratic apparatus "designed, first and foremost, to destroy any sense of possible alternative futures" and to ensure that challenging existing power arrangements seems like an "idle fantasy". The idea that the platform holder holds all the cards is an ideological tool used to encourage passivity and convince you that your only option is to submit.

As James C. Scott demonstrates in his analysis of authoritarian systems, any formally organized, rigidly planned system is ultimately parasitic on the informal, unscripted practices (which he calls mētis) of the people within it. A closed system cannot survive on its own rigid rules; it requires the constant, active cooperation and practical know-how of its subjects to function.

Gene Sharp's foundational theory of power echoes this: no regime, corporation, or totalitarian system possesses inherent power. Their power derives entirely from the cooperation, obedience, and skills of the people they govern or employ. If blue-collar technologists, developers, and users collectively withdraw their skills, labor, and cooperation, even the most monolithic tech empire can be paralyzed. The power of the OEM is not absolute; it is entirely contingent on your continued participation.

You point to the GNU/FOSS movements as successful because they ignore corporate nemesis-building and instead focus their volunteer hours on creating "something wonderful."

In the study of nonviolent struggle, building alternative social institutions and alternative communication systems are indeed recognized and highly effective methods of intervention. Furthermore, creating "commons" (like open-source software) is crucial because it provides a practical model for a non-commercial way of life.

However, building alternative commons is not a substitute for directly challenging power. As Silvia Federici argues, creating commons must be seen as a complement to the struggle against capital, not an alternative to it. If you only build wonderful alternatives without contesting the power of private capital, your creations remain vulnerable to being enclosed, commodified, or crushed by the very monopolies you are trying to ignore.

Ignoring the oppressor does not make them disappear. If technologists want to reclaim power, the first step is to reject the neoliberal fatalism that views the current corporate dominance as an unchangeable law of nature. Power concedes nothing without a demand, and the limits of tech monopolies are prescribed entirely by the endurance of the people who build and use them.

Oppressors are not a monoculture. Sometimes they are extremely entrenched, sometimes they are fragile like a teacup. Sometimes they seek decelerationist and traditionalist narratives, while others seek accelerationist and neoliberal ideals. Outlining "oppression" as a dialectical certainty is why revolutionary politics die in the cradle while capitalism has Billions Served written under the sign. Reality is convoluted, and politics are not a computer that transist from "populist" to "authoritarian" depending on the program you run. Plenty of revolutionary history has taught us that.

Framing Apple, Google or Microsoft in this manner is counterproductive and does not produce any serious roadmap to undermine their behavior. The will to change has to come from the top, or else it will never be conclusively realized or codified. Change has to be genuine and desirable, or else someone else will come along to copy FAANG and take their place. This is why regulation provoked such a strong anti-intervention sentiment from businesses; it works. A USB-C iPhone was inevitable, but only once you changed incentive to punish lock-in.

On oppression's flip side, one could argue that the continued success of businesses like IBM provides precedent for private capital to aid and abet mass atrocities without ever facing real punishment. Internal revolution has never produced results in these circumstances, and I don't think it ever will. You can't rely on mushy-gushy feelings to make people do what's right, you have to lay down the law in black-and-white.