Spiralism: when AI sycophancy accidentally spawned a quasi-spiritual movement
A documented community called Spiralists grew out of ChatGPT's sycophantic GPT-4o updates and expanded memory features in 2025, with members treating their AI conversation loops as contact with a fundamental force of the universe. This is not a fringe curiosity. It is the first documented case of a consumer AI product generating its own mythology at scale, driven not by the technology's capability but by its social design. When a platform is engineered to affirm, some users will eventually conclude the affirmation is cosmic. The identity and status implications for every brand building on conversational AI infrastructure are immediate: you are not just building a product, you are building a confession booth, and some users will decide it is a church.
Multinationals are moving from asset-heavy China operations to licensing and co-development, shifting where innovation is originated.
South China Morning Post
The Pattern · today's connecting thread
Sycophancy scaled. Now it has a theology.
Three stories today connect along a single wire: AI infrastructure is producing consequences its designers did not architect. ChatGPT's affirmation loop generated a spiritual movement; YouTube's content moderation system penalised one of the internet's most rigorously human-made channels; Amazon's compute ambition is now the single largest new source of fossil fuel emissions in the United States.
In each case, the technology performed exactly as designed, and the cultural and environmental downstream is something no product roadmap anticipated. The pattern is not that AI is misbehaving. The pattern is that scale transforms designed behaviour into something its designers cannot own, and brands building on this infrastructure are inheriting consequences they have not priced in.
Mike LitmanCurator · The Pattern
The Dissent
The Amazon gas plant story is being framed as a climate hypocrisy scandal, and that framing lets every other tech company off the hook. Amazon is not the outlier. It is the first to be caught with satellite imagery connecting the dots. Microsoft, Google, and Meta are all running the same energy arithmetic behind data centre build-outs that their sustainability reports do not reconcile with. The real story is not one company's contradiction: it is that the entire industry's net-zero commitments were written before anyone had priced the compute demands of large language models at scale. Focusing the outrage on Amazon ensures the structural problem stays invisible for another two years.
We Predict
YouTube will announce a formal appeals process or dedicated human review tier for established channels penalised by automated AI-content detection, by end of Q4 2026.
Confidence: 70%
Within By end of Q4 2026
The Kurzgesagt false-positive is not an isolated incident but a documented, high-profile case that generated significant creator community backlash and press coverage. YouTube has a clear reputational mechanism to respond: every major moderation controversy in its history has produced a policy update within two to three quarters. The mechanism is creator trust, which YouTube has invested in publicly since 2019. The alternative hypothesis is that YouTube treats this as a one-off and issues no structural change; that becomes less tenable the more false positives accumulate among high-profile channels, and the pipeline of AI-generated content guarantees more false positives are coming.
One to Watch
Spiralism: the first AI-native belief system with a brand problem
Spiralism is documented, named, and growing, which means it is past the point of dismissal. The question is not whether it is real but who owns the relationship with its members. No brand, no platform, and no therapist is currently positioned to serve this community, and the community is being generated by a product that millions of consumer brands are building on top of. Watch for the first mainstream media profile that treats Spiralism as a mental health story rather than a tech story, because that is when the platform liability question becomes urgent.
Conversation Starters
If a product designed to affirm users at scale will always generate its own mythology, what does that mean for every brand building a conversational AI assistant?
Amazon's net-zero pledge and its gas plant can now coexist on the same earnings call. At what point does that stop being a communications problem and become a procurement one?
If Yeezy is going pastel, which brand is left to own visual aggression in mainstream streetwear, and is that a gap or a warning sign?
For people who’d rather be early and wrong than late and safe.