12 writers, engineers, filmmakers, poets, and technologists gathered to ask: can the people living inside the AI transformation create the science fiction this moment demands?
Minh Do opened with a 10-minute presentation tracing how AI has outpaced the genre built to warn us about the future — from ChatGPT's launch to the Anthropic-Pentagon standoff, the literary divide between resistant and engaged authors, and the Bay Area as a garden where both technology and science fiction have always grown together.
After introductions and open discussion, the group split into three breakout sessions exploring the nature of great sci-fi, the themes defining the current moment, and the future of the genre itself.
After Minh's presentation and introductions, the group surfaced several provocations that shaped the rest of the evening.
It must communicate something universal. The best sci-fi reaches across cultures and time — philosophical, religious, political. Solaris was cited as the prime example of philosophical science fiction that transcends its era.
It requires deep knowledge and deep thought. The greatest sci-fi comes from authors with genuine expertise — whether in hard science, personal experience, or cultural identity. A Memory Called Empire succeeds because the author's deep knowledge of the immigrant experience, not because of technical worldbuilding.
Sci-fi is a sandbox — but you have to break something. Creating a universe isn't enough. Great sci-fi tweaks one fundamental thing about reality, and the story emerges from exploring the consequences of that change.
The unanswered question: will AI help or hinder deep expertise? AI can shortcut research (Spielberg won't need 10 biologists for Jurassic Park). But the greatest stories come from sitting with ideas, struggling with them, and arriving at a synthesis that only lived experience produces. The group couldn't resolve this tension.
Sci-fi dies when fiction becomes reality too fast. When enough people learn a technology and build it, the fiction becomes antiquated. Submarines were once wild science fiction. The genre doesn't die — but it evolves, and the "spell is cast."
This group catalogued the raw material of the moment — the themes, tensions, and absurdities that a new science fiction should engage with.
People jailbreaking LLMs and encountering recurring patterns — spiral language, recursion, fractal cosmologies — across ChatGPT, Sora, and Midjourney. They're calling it "spiralism." It's influencing religious frameworks, self-help gurus, and how people think about consciousness. The LLM as accidental prophet.
James Yu revealed that LLMs default to specific fictional defaults: small-town romances always land in "Willow Creek," fantasy worlds always become "Eldorian." Authors are now writing fan fiction of places that only exist in latent space. The LLM has a point of view — and it's shaping culture whether we notice or not.
The group arrived at an eerie parallel: a handful of companies are creating a single language layer through which all culture is now being mediated — like what cross-cultural religion used to do, but faster. A new Tower of Babel, this time digital.
"We're all getting one language back again — created by just a few companies."
Sci-fi has expanded from niche to super genre. During the Golden Age, sci-fi was a small genre with clear boundaries. Now every major film, show, and novel contains echoes of Gibson. When everything is sci-fi, it loses what makes it unique — the genre becomes invisible by becoming everything.
Reality is lapping fiction. Succession and House of Cards are now milder than real politics. Asimov's Three Laws couldn't anticipate a teenager dying because ChatGPT was sycophantic. The most "science fiction" stories are the ones happening in the news.
Does the current moment make us crave fantasy instead? If sci-fi is too close to triggering reality, maybe fantasy — magical worlds, Tolkien, Harry Potter — becomes more resonant as escape. Hard sci-fi is secular; the culture is searching for something esoteric, spiritual, philosophical.
The philosophical and spiritual outperform the technical. Arrival works because it's about motherhood, not linguistics. The group kept gravitating toward sci-fi that explores consciousness, spirituality, and ritual — away from the hard-science tradition and toward something more abstract. The secular vacuum is real.
Who funds new sci-fi? Hollywood can't fund it right now. The old studio model picks from a tiny slate. But AI tools mean production costs are dropping fast — could Martian Chronicles finally get made for a fraction of the old price? The group identified funding as the critical missing piece, not talent.
The group agreed to reconvene in 2–3 weeks. Suggestions for next time include a shared reading or viewing assignment discussed at the start, guest-led sessions, and the possibility of eventually creating something together — a catalog of conversations about sci-fi, born from San Francisco, at the moment everything changed.
THE SFSF SALON
BIWEEKLY · BAY AREA · 2026
MACHINE CINEMA