The Tragedies of AI May Precipitate an Age of Non-Conformity

Source: MSW@USC Diversity Toolkit: Guide to Discussing Identity, Power and Privilege

We should worry “boundaries” increasingly set by algorithms will mean people may lose authorization to operate outside the “authenticity” boxes they’re placed in by others; denial of identity freedom may require generating more sophisticated forms of non-conformity.

In a couple recent blog posts I point out how an American expansion westward was driven by slavery economics and marked by concentration camps and genocidal campaigns.

Victims of humanitarian disasters in America were in theory offered an “exit” from total annihilation, if they chose to conform entirely by abandoning any freedoms of identity. Expression in speech wasn’t harshly limited, you see, just an “authenticity” of identity.

The Native American people were cruelly forced to operate on soon-to-be industrialized U.S. platforms, which meant sacrifice of self-determination. They were told at gunpoint to dress differently, speak a different language, sing/play/dance to different music, eat different foods, cook different meals, do different work…they were forcibly transformed from private owners to public “users” and every aspect of their identity had to change to conform to the encroaching immoral platform owners.

This has aptly been described as “being spoon-fed out of a bucket of whiteness every day.”

The “longhair” revolution of the 1960s often attributed to white “freak flag” communities, for perspective, was in fact adopted from a Native American movement to reclaim their identity rights.

…the physical cutting of hair is a manifestation of the loss of a loved one, a loss of a relationship, and a loss of a part of self…

You perhaps can see why identity self-determination was so important. Native Americans suffered greatly under U.S. tactics that forced them to conform to “Christian” identity requirements or face starvation (a 1902 Bureau of Indian Affairs “haircut order” required short hair to receive rations).

We unfortunately, despite lessons from the past, see a similar conformance campaign ethic driven by Facebook today. While being implicated in genocide, Facebook has taken a tactic to harshly prosecute people for freedoms of identity instead of clamping down on the speech that actually foments genocide.

Anyone attempting anything less than what platform owners consider “real” or “authentic” is eliminated from the platform. Facebook mistakenly calls this culling of identity freedoms a security measure, which to me reads like someone studied only the imitation game (Turing test) in computer science and skipped history classes.

Meanwhile Facebook does very little or nothing at all to address the real harms caused by the people they judge as authentic.

…we will not send organic content or ads from politicians to our third-party fact-checking partners for review.

It’s like Facebook saying the white supremacist diatribes of user Stanford couldn’t be blocked from facilitating genocide on their platform because he did so from an authentic identity. Only if he had done things like put on a strange coat of feathers and wore long hair, or grew a beard and put on a hat to look like the user Lincoln could he have been de-platformed.

From both a security and history perspective, Facebook has been wrong to blindly repeat the worst mistakes in history and force a dangerous conformity on their self-serving expansionist terms.

Edge cases of true impersonation (an integrity risk, such as stolen valor and authorization fraud) exist and should be stopped. “Deep fakes”, to that end, has been generating a lot of excitement. Yet it mostly begs old questions about whether new low-cost generation of content still should be regulated as art or expression.

People need to consider seriously whether a much greater threat to freedom is the opposite effect, Facebook operating a Kafkaesque identity conformance program of “deep realism” (e.g. already for several years I have met with government regulators concerned about harms to society including national security and the economy).

The risks from identity abuse edge cases of fraud/authorization are far less compared to dangers of militant removal of freedom and creativity of identity on global platforms. One could even argue, for example, the entire concept of the modern cosmopolitan lifestyle made famous in markets like NYC, Paris, London…is the high art of regionally managed platform identity freedom.

An appropriate reaction to centralization power for platform conformity pressures from a genocide-facilitating for-profit company like Facebook perhaps will be a non-conformity revolution.

The best intelligence analysts already know this tactic. Adversaries love conformity because they can predict moves so easily and camouflage isn’t even necessary. A “weird” analyst by comparison becomes a nightmare of any adversary, because prediction of how they will react becomes impossible.

Another appropriate reaction is rotating focus back to harm, which means pushing a standard for filtering by actual risks while letting people express themselves from whatever identity they choose to develop and the communities to which they belong.

Take for just one example the concept of gendered color coding.

Pink is considered by some even to this day a shade of the “warlike” red, as in the British Red Coats. It stems (pun not intended) from Oliver Cromwell’s “New Model Army” adopting distinctive “Tudor Rose” dye as their uniform for war.

This of course was reconsidered around WWI when machine guns and snipers were killing anything identifiable, and military uniforms shifted to more muted tones to impersonate surroundings.

Meanwhile others, particularly in America, somehow came to see pink as their tool for encoding female identity:

“…a kind of early gender coding that worked especially on young girls. The decade of the Fifties was characterized by an ideological emphasis on conformity, and by fashion images that were sharply age- and gender-specific.”

In that sense pink in America really represents strangely planned attempts to make science more anti-social and eliminate women who scored high on social-good measures:

…programming’s shift from a women-friendly occupation to one that is hostile to women. In the 1950s and ‘60s, employers began relying on aptitude tests and personality profiles that weeded out women by prioritizing stereotypically masculine traits and, increasingly, antisocialness.

Even with this history of encoding and sexism it should never be wrong for any gender to associate pink with their identity. There generally is no harm of the color (with rare exceptions, such as war). Compare that to someone wearing an offensive hate speech patch or logo designed to do harm and expand suffering, on the other hand, and you see more logical security control areas.

Facebook’s genocide-facilitating platform is likely coming for your community with its AI, trying to get a lock on all the identities using things like “coding that worked especially on young girls”.

Thus more non-conforming behaviors should develop as fundamental survival tactics. Where can your data live that it will have freedom for identity?

I have long advocated for, and concur with latest research, that we need to assess code as potentially malicious (whether human or machine) and emphasize filters as a useful control for individuals to operate.

Researchers at the University of Cambridge have proposed a software program that treats online “hate speech” like a computer virus.

Scientific anti-pollutant concepts of continuous collection for known harms are a far more sensible approach than those of Facebook (repeating mistakes of American history) pushing to oppress identities through opaquely self-serving and centrally-planned concepts of what they call authenticity.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.