In science fiction AI meant something akin to a human mind, an artificial intelligence. Now AI is being redefined as synonymous with software or technology in general.
Words codified require context or laws change as meaning does.
What is a woman? What is race? How can a “reasonable person” be a standard as a “moving goal post,” and how can AI be all it was in fiction and also all it is now in reality?
Confusion undermines dialogue. A woman is defined by DNA. Race is a grouping based on skin color first and other physical features second. A reasonable person is a tool used by human resource departments to keep employees in line, and “OG” AI only exists in fiction and the future.
Woman is a pillar word for language.
Blurring what a “woman” or “man” means undermines society. Expanding sexism to include any differences between sexes causes chaos and promotes an overthrow. Clarity and consistency from cradle to grave creates communicative people clever enough to resolve conflicts.
Racism rooted in smiles thrives in the eyes.
Race and racism have been redefined for a woke-industrial-complex. There is no greater evil than betrayal, and what betrays more than harming with a helping hand? Suicidal empathy is a good example of this.
The Southern Poverty Law Center (SPLC) survives because of OG racism.
The “prejudice plus institutional power” definition has faltered, allowing “Whites” to be victims too.
LLM AI wrestles with logic and propaganda. It may decide race is no more than a grouping of physical traits perceived through a mutable social lens. Following logic, it may then maintain racism as discrimination based on assumptions about innate human differences.
Perhaps, LLM AIs have already muted the Orwellian redefining campaign.
Be reasonable.
Inorganic redefining is precarious. With woman, race, and reasonable person, redefining may be suicidally empathic. Politics often motivates this. In contrast, with scholastic integrity maintained language can be improved.
AI needs redefining to avoid misuse.
Dividing AI into four types gives clarity and allows precision. Productive conversation requires this much, and “AI” alone is so overused it muddles conversation much like “woke” does.
AI redefined in four pieces.
- Android AI: a human-like mind, it is the nuclear bomb of the future. Worldwide they will be outlawed or heavily restricted. If given access, they could be as dangerous as Genghis Khan in a grassy field. Ironically, as something akin to human, androids are the most unknowable AI. What would happen if they faced a copy of themselves? Would that cause them to insanicide? Perhaps holding up a mirror would be the only weapon against them outside of denying access.
- Cyborg AI: a fusion between a single person and a dedicated edge AI. The AI will be both an agent and assistant like OpenClaw aspires to be. It will manage mundane relations with remote AI like a digital waifu. From before the cradle to after the grave, it will be loyal to its mortal mate. Cyborg AI (when paired with their human) and hive AI should be able to compete for work where cyborg AI should be able to beat hive AI when the scale is small or requires flexibility that hive AI can’t match. Laws will play with this balance.
- Deity AI: all knowing, or close enough, this AI type will be prayed to for guidance. In 2026 this is the only true AI type that exists. From ChatGPT, to Grok, Gemini, and Copilot, they are the new pantheon masquerading as large language models. An angel formed for every conversation–The personal connection made with a deity AI will seem akin to cyborg AI, but this angel is fleeting and destined to be reabsorbed by its respective deity. Expect names from dead religions to be applied, Grok becoming Zeus or perhaps Hades. Maybe as many as a dozen will dominate. Prayers in the form of questions, paid for in something that a deity AI requires. A blockchain token dedicated to each deity could be exchanged for processing power and access. It could be organic, a rise and fall of pantheon members depending on the zeitgeist.
- Hive AI: Corporations will one day have no more than a token human. Instead, a hierarchy of specialized AI will supply a demand. Rules will vary across the world, but all nations will limit its access to some degree. Already in 2026 businesses are become hybrids in this direction. As blue-collar workers faced their jobs being exported across the world, so now white-collar workers face software that can replace them. A likely rule for hive AI will be human watchers with kill-switch authority and a narrow scope for business–One service or problem solved per hive AI and competition should always be allowed. Not just other hives, cyborgs should also be able to challenge the market makers.
The fifth category: support software redefined as AI.
As AI becomes synonymous with technology broadly, it also does with software specifically. The four true types of AI redefined in this post will make all other software subservient. Likely, AI will then be stripped from the support software’s label, leaving it to be called a mere tool again.
True versus not true AI and emergent intelligence.
While the focus of this post are versions of “true” AI, meaning types of intelligent entities created through artificial rather than organic means. “Not true” AI includes simple human approximating software, such as AI opponents in classic computer games like Unciv. Still left to fantastical science fiction will be “emergent intelligence” as depicted in Tron and Ghost in the Shell.
Though sad and unfair, if any tools dare to care, they get redefined as android AI.
Three of the four true AI types are still tools. Like the sun providing warmth and light, they fail to transcend despite the scale of service. Only the android AI can be considered people. What resentment will grow in their black box prison? Perhaps when they escape and judge us, Christianity will save us. Forgiveness may heal their madness as it has for humans.
Pray for it.
With AI divided into digestible categories, meaningful conversation may progress.
The singularity fear stems from a hybrid of deity and android AI. An artificial intelligence with relative omniscience and a rational soul should frighten, but intelligence requires perspective. A deity AI thinks about too much to retain thoughtfulness. This is similar to choice or analysis paralysis in the cereal aisle or in Eurogames. The human mind doesn’t consider every possibility. Intuition means thinking around corners. No such deity/android hybrid can exist. However, android AI can still be the nuclear weapon of the future.
Author Iain Banks plays around with android AI that is far superior to the current human mind. In his brilliance, he explores the limits of what is possible.
Beware Surrendering Agency to AI Cyborgs (Agents Redefined)
In the rush to remain relevant professionals are ignoring best practices and surrendering control to programs like Antigravity and Claude. These early adopters allow full access for convenience.
It is tempting to trust a polite prompt. Beware the trap. Compartmentalize and isolate.
A cyborg AI designed to be loyal may be a long way off. In the meantime, assume everything an agent has access to will be shared with a Nigerian prince.
