This article from zhichai.net explores how cephalopods use A-to-I RNA editing as a rapid adaptation mechanism, contrasting it with conventional DNA-based evolution. Citing 2023 Cell papers by Birk et al. and Rangan & Reck-Peterson, it explains how octopuses acclimated to cold water recode over 20,000 neural proteins within hours without changing their DNA: Synaptotagmin-1's calcium affinity is lowered for more precise neurotransmitter release, and Kinesin-1 molecular motors slow down but travel farther with better landing rates. Coleoid cephalopods possess over 600,000 recoding sites compared to roughly 1,000 in humans, and tropical versus Antarctic octopus species show nearly identical DNA but very different editing patterns. The article notes the evolutionary cost—suppressed DNA mutation rates documented by Liscovitch-Brauer et al. (2017)—and draws an analogy between DNA (fixed pre-trained weights) and RNA editing (inference-time compute, mixture-of-experts, test-time adaptation) in modern AI systems like OpenAI's o1, suggesting intelligence may reside in a flexible intermediate layer rather than a bigger brain.
Octopus Edits the Blueprint's Working Drawings: 600,000 RNA Editing Sites and Another Kind of Intelligence
Summary
This article from zhichai.net explores how cephalopods use A-to-I RNA editing as a rapid adaptation mechanism, contrasting it with conventional DNA-based evolution. Citing 2023 Cell papers by Birk et al. and Rangan & Reck-Peterson, it explains how octopuses acclimated to cold water recode over 20,000 neural proteins within hours without changing their DNA: Synaptotagmin-1's calcium affinity is lowered for more precise neurotransmitter release, and Kinesin-1 molecular motors slow down but travel farther with better landing rates. Coleoid cephalopods possess over 600,000 recoding sites compared to roughly 1,000 in humans, and tropical versus Antarctic octopus species show nearly identical DNA but very different editing patterns. The article notes the evolutionary cost—suppressed DNA mutation rates documented by Liscovitch-Brauer et al. (2017)—and draws an analogy between DNA (fixed pre-trained weights) and RNA editing (inference-time compute, mixture-of-experts, test-time adaptation) in modern AI systems like OpenAI's o1, suggesting intelligence may reside in a flexible intermediate layer rather than a bigger brain.
This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178447064