• Notes
  • Revised to be more concise. This is a statement of opposition, not an invitation for debate. I believe the harms are self-evident. My positions are unmoveable except through my own revising.
  • My opposition is on philisophical, ethical, and moral grounds. Utility is tangential and questionable and ultimately not important.
• Generative AI / LLMs
  • I consider myself a conscientious objector. Full opposition, without reservation. Outside of exploited workers, I condemn these systems and their users.
  • It is not a tool. A paintbrush does not generate an entire completed visual work. A hammer doesn't build a chair by being swung.
  • It does not think. It is not intelligent. It mimicks speech patterns at best. It is a facade.
  • It does not democratize. It commodifies under the guise of democratization. It flattens and removes the process. It does not provide opportunities for someone, it replaces them and their efforts.
  • It is unethically trained. Data is unethically obtained. Web servers are attacked by crawlers. Crawlers specficially are tailored to appear as other agents to purposely obtain data against explicit structures of refusal.
  • It is a slot machine with an undeniable emotional and psychological pull and users start rationalizing the same way an addict would.
  • Erodes critical and creative thinking.
  • It is beginning to destroy open-source.
  • Its anthropomorphization carries with it real and (tw: suicide) already harmful dangers of many kinds- of psychosis, misplaced trust and affection, and other bonds. It becomes a companion for some, a drug for many, a god for others, none of which positions it should be allowed to have.
  • Arguably one of my most important points here: prompting is not expressing, it is commissioning. People prompting generative AI are not artists, coders, writers, etc. They are frauds that have chosen to commission a system rather than a person.
  • It is an architecture of subsumption and a threat to privacy, both from non-consensual data collection and from convincing users to freely give these systems very sensitive information.
  • It is a brazen attempt to generate revenue in circular patterns between companies to make the rich richer at the expense of everything else.
  • Its use creates dependency. Productivity dependency at an employment level and personal dependency at a deskilling level.
  • It is an intentional threat to worker power. Its use creates a burden of work for others. Sows distrust. Changes productivity expectations. Damages workplace dynamics.
  • It is a grift/fraud enabler. It saturates every field it touches with low-quality high-volume fake work by skill-free people that want praise or status without having earned it.
  • It is notable to me that the defenders and owners of generative AI often are the same people that do not value art or the process of making it and instead see it as a burden on the artist.
  • It is not simply just a technological advancement like those of the past. It is a paradigm shift that signals a finality of removal of humanity from all manner of work, whether creative or otherwise.
  • The Luddites were not anti-tech, they were right, and propaganda has tried to cover up that fact. We need this movement again now.
  • Inevitability is never an argument in favor of something. It is at best an argument for having some kind of response to it. There is no obligation to assimiliate or be defeatist and unnecessarily pre-compliant even if something feels here to stay.
  • Opposition should be adversarial in nature, not just rhetorical. Make it socially shameful and costly to use and especially to bring it into spaces meant for humans. Out of solidarity, the only exception to be made here is for workers that are forced to do it. However, I call on workers to do what they can to make its use undesirable within their companies. Erode their power from the inside out.
• Other criticisms and ethical concerns
  • Against the anthropomorphism of any AI. They are not and will never be humans and they should never act like them.
  • Against agentic AI. They do not have human understanding, emotions, ethical reasoning, and so on and thus should not act on behalf of them.
  • The apocalyptic big scary sentient AGI or Super Intelligence is pro-industry marketing masquerading as criticism. It pre-supposes an intelligence these systems do not have (despite the marketing pushed by their owners). Be suspicious of anyone speaking supposed criticism primarily from this angle.
  • It's environmentally destructive and furthers environmental racism/classism. Single-query energy assessment is a deflection. It's no secret that these systems and their owners are using massive amounts of power and water and openly declare desire for plants to power them.
  • Against algorithmic discrimination and computers as scapegoats for responsibility, and the dissolution of human checks, balances, and watchdogs that comes along with that.
  • Output should not be given copyright, freedom of speech, or press rights
  • Fully oppose all use in warfare, surveillance, incarceration or other apparatuses of destruction, power, and policing.
  • Fully oppose all use in livelihood determinations (financial eligibility, life/death/medical, assistance, benefits, etc).
• What about non-LLM AI?
  • If you have a simple markov-chain-based bot for humor purposes, that is only trained on a very small set of output local to its location, is very obviously a bot (labeled and unconvincing in its output) and it uses minimal energy, and so on- it's really not a threat.
  • Non-LLM narrowly-tailored deep/machine learning trained on only ethically-sourced data, built for a specific positive purpose (folding proteins, etc) is less of a concern, though not without flaws, suspicious marketing, and so on. Pays to remain vigilant but there can be okay use cases here.
  • Even narrowly scoped uses should be human-guided and checked by individuals with enough knowledge and skill to do so
  • The biases in the input/output should be considered and intently looked for
  • The humanity of those involved and those affected by the eventual output should be a top consideration
  • Other ethical concerns in above sections also apply here

Exceptional posts on AI worth reading

• Tackling the supposed use of AI for disability and disability-washing of AI
  • I Will Luddite AI - Sightless Scribbles
  • Q&A: AI and Disability - Dan McQuillan
• Other topics
  • The rise of Whatever - Eev.ee
  • On Vibe Coding - Tante.cc
  • AI As Class Warfare - New Intl Mag
  • Drew DeVault: AI Crap