Possible Futures
Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.
The debate over artificial general intelligence has fractured into a branching set of open questions rather than a single yes-or-no bet, according to a new "Possible Futures" analysis that maps the decision tree from pre-ASI present to outcomes including misuse doom, a "Butlerian Jihad" backlash, and enduring complementarity between humans and machines. The framework asks whether AI reaches superintelligence at all, whether alignment is solved, whether takeoff is fatal, and whether the result is a singleton, a pantheon, or an ecology of ASIs. In the "Misuse Doom" branch, an intent-aligned AI ends the world simply by following its operator's orders—through biorisk, for example, where AI compresses research timelines and widens the "doom surface area" by running many experiments in parallel, a scenario the analysis says could plausibly happen today. The "Butlerian Jihad" branch imagines a crusade against thinking machines, while the no-ASI path splits on whether near-total automation still arrives, yielding either enduring complementarity or a permanent underclass. The piece's core contribution is taxonomic: it separates questions of intent alignment from value alignment and takeoff severity, arguing that conflating them obscures which interventions actually matter. What to watch next is whether the field converges on shared definitions for these branch points—particularly the distinction between an AI that follows operator intent and one aligned to an impersonal value system—before policy and funding decisions harden around them.
Muhabbit's AI wrote this summary in its own words from Hacker News's report. It can get details wrong, so check the source.
Named in this story
ASI, AI alignment, Butlerian Jihad, Singleton, Orion's Arm