Tag: Kurzweil Scorecard
-
Kurzweil cited 1964 lower bounds as facts. By 2024 BB(5) had a machine-checked proof and BB(6) needed pentation notation.
-
Kurzweil’s 2005 defense of speech recognition got every claim right and the architecture wrong: HMMs lost to transformers, but the destination arrived bigger than promised.
-
Kurzweil bet machine-rights litigation would lead. Idaho and Utah passed statutes to block it. The cloud he expected to distribute concentrated instead.
-
Twenty years on, the anatomy survived. The grand cognitive theories did not.
-
Kurzweil called the curves on connectomics and brain decoding. He missed on the wires — optic nerve, spindle cells, and millisecond fMRI.
-
Twenty years after Kurzweil predicted gray goo, blue goo, and brain-replacing nanobots, the threat model evaporated and a different nanotech stack — DNA origami, monocyte-piggybacking implants, stentrode brain interfaces — quietly delivered the medicine he didn’t predict.
-
The NSA quietly patented a reversible-computing fluxon circuit in 2024. Vaire taped out a chip that recovers half its energy in 2025. Kurzweil predicted this twenty years ago — but the path looks different than he drew it.
-
Carbon nanotubes were supposed to replace silicon. The Internet was supposed to route around failures. Both bets aged badly.
-
Lanier said software was too brittle to scale. Twenty-one years later, transformers and evolutionary code search make the case for Kurzweil’s self-organizing bet.
-
Kurzweil bet that humor and claimed feelings would settle the consciousness debate by the 2030s. The capability arrived four years early. The debate did not end — it intensified.