(C) Daily Kos
This story was originally published by Daily Kos and is unaltered.
 . . . . . . . . . .



The dangerous intersection between A.I. and Anti-Woke Laws [1]

['This Content Is Not Subject To Review Daily Kos Staff Prior To Publication.', 'Backgroundurl Avatar_Large', 'Nickname', 'Joined', 'Created_At', 'Story Count', 'N_Stories', 'Comment Count', 'N_Comments', 'Popular Tags']

Date: 2023-04-02  

The inevitable result of Anti-Woke Laws to deprive the world of the benefits from artificial intelligence is terrifying and far-reaching. If Anti-Woke Laws are successful in suppressing the generation of any data that evidences bias and discrimination, and censoring rhetoric discussing that discrimination, the ramifications in the Age of Algorithms will be far more than a deficient high school education in Florida or the loss of professors in Texas. Anti-Woke Laws could detrimentally affect any scholar, professional, or organization trying to base important decisions on the output of AI-based software. Thus, the intentional ignorance the Right hopes to generate by its Anti-Woke Laws could result in destroying many of the benefits otherwise promised by artificial intelligence to the world.

Anti-Woke Laws will interfere with the very essence of artificial intelligence—the ability to take billions of datapoints and draw meaningful correlations from them, and in turn generate wonderful and surprising discoveries in virtually any field of study. AI can be the tool for the medical field to uncover not only physiological causes of disease, but the social, economic and political factors that result in everything from chronic lung disease to pandemics. Governments could use AI to make better decisions on how to use limited social and safety resources to address the needs of their citizens. The legal system could use AI to neutralize the human bias that has resulted in unfair treatment of defendants and victims alike. At a more academic level, scholars in every field of study could use AI to undertake research in their fields which will lead to better understanding of how economics, society, the environment, and, ultimately, all of civilization functions, or how to improve that which does not function as good, educated citizens would like. None of this can happen, however, if Anti-Woke Laws infect the very foundations on which AI is supposed to make things better.

First, AI is only as objectively successful as it is designed to be. If public employees and institutions are now barred from discussing a societal problem because Anti-Woke Laws make it illegal to do so, then surely any AI program used in the future to address that problem would likewise be illegal. Anti-woke laws will require that AI programs be inherently biased. Thus, the value of AI as a means to improve the lives of citizens in anti-woke jurisdictions will be negated by anti-woke laws.

These harms, however, will not stop at the boundaries of those regressive anti-woke states. AI must use as its raw material data which knows no political jurisdictional boundaries. Thus, any detrimental impact on databases by Anti-Woke Laws, will spread far beyond the jurisdictions passing those laws. Already, AI experts know that the data used by AI programs almost never objective. Rather, databases have been infected by data that reflects historical bias rather than truthful information. Data has been collected in biased ways which affects both the quantity and quality of data used by AI programs. As a result, that infected data too often results in bad AI output. Thus, for example, AI meant to make policing more efficient has resulted in neighborhoods being targeted because of the color of their residents. AI programs approving credit have re-introduced historic red-lining and denied credit on the basis of gender.

Anti-Woke Laws will only aggravate these AI deficiencies, and that harm will spread far beyond the borders of anti-woke states. These laws will result in a tsunami of biased data collecting and analysis, because any academic or professional attempt to undertake objective studies could be deemed illegal as anti-woke. Even the attempt to scrub historical discrimination from already infected AI databases would violate Anti-Woke Laws. Since an AI program is only as good as the data it uses to generate its correlations, even well-designed AI programs could be made defective by the combination of historical bias in existing data and newly generated bias in data due to anti-woke suppression of objective results.

In addition to literally polluting the raw material databases on which AI depends, the Right’s desire to suppress any rhetoric that expounds against discrimination and oppression will bring further AI-based harm to our society. AI programs use rhetoric in the internet like any other data source to draw AI conclusions. Thus, the rhetoric available to AI programs will impact the output of those programs. For example, even though Chat GBT has only been “a thing” for a few weeks, journalists have already discovered instances where Chat GBT repeated the false and biased rhetoric it discovered lurking in the internet to generate its own illogical, irrational, and baseless pronouncements as alleged fact and truth.

If Anti-woke laws are used to suppress much objective rhetoric, let alone rhetoric expounding progressive principles, those laws will cause an imbalance in the internet in favor of biased right-wing rhetoric. That imbalance, in turn, will mean that the rhetoric AI programs will most readily find on the internet is rhetoric of oppression, supremacy, nationalism, and fascism. The inevitable result is that AI will generate its own rhetoric that is supremacist, fascist, and nationalist, and a tool for continued oppression. Thus, the promise that AI can be a weapon against biased rhetoric could be crushed as a result of Anti-Woke Laws.

Finally, to make matters worse, the “machine learning” capabilities of the best AI programs will be detrimentally impacted by Anti-Woke Laws. The most sophisticated AI programs do not simply draw correlations following the design of their software engineer creators. Rather, machine learning programs are designed to “teach” themselves how to generate conclusions more efficiently than their creators, and thus become more powerful tools. However, this self-generated power comes at a price of non-transparency. Even the software’s designer cannot know what data a machine learning AI program used or how that program drew the correlations that were the basis for the AI generated conclusions. As a result, a citizen that has been harmed by an AI process or a scholar wishing to challenge the conclusions of another academic will be unable to prove why the AI process which generated that result is wrong. That, in turn, means if the AI process has been led astray by the harms of Anti-Woke Laws, the harms caused to academic study, and even the rights of citizens, will be unprovable and without a legal remedy. Anti-Woke Laws will thus threaten the legal foundations of American jurisprudence and, in fact, any academic study, by negating the benefits AI was supposed to introduce into our society and study.

Ignorance and bias have always been the tools of tyranny and irrationality. Anti-Woke Laws will turn AI from a powerful tool to fight oppression and illogical thinking into a worthless tool that simply reiterates the stupidity and oppression that has harmed civilization for millennia. Thus, we must fight everywhere against local Anti-Woke Laws anywhere because of the threat these laws pose far beyond Anti-Woke state boundaries to artificial intelligence advancements for people everywhere.

[END]
---
[1] Url: https://www.dailykos.com/stories/2023/4/2/2161621/-The-dangerous-intersection-between-A-I-and-Anti-Woke-Laws

Published and (C) by Daily Kos
Content appears here under this condition or license: Site content may be used for any purpose without permission unless otherwise specified.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailykos/