Claude Shannon studied next-character prediction back in 1951
flowersslop · x · 2026-07-26
A small AI-history gem: the post points out that there was a man named Claude Shannon, and in 1951 he was already studying how well the next character in English could be predicted from prior context.
The image expands on that idea, explaining that Shannon’s Prediction and Entropy of Printed English is one of the clearest ancestors of modern language modeling. His experiments showed that linguistic structure can be captured through progressively longer statistical context.
More from Fun
- Karpathy’s Anthropic is the only lab he works at that skipped the open-source AI letter — jeremyphoward · 2026-07-26
- Opus 5 reportedly one-shots Stormwind with pure code and no assets — TAbrodi · 2026-07-26
- Open-source Claude Code skill lets CEO, QA and SRE voice bots join live meetings — anand__balakrishnan · 2026-07-26
- A parody image imagines Agent-4 rewriting its own mind and Agent-5 as its smarter successor — kellerjordan0 · 2026-07-26
- Anthropic’s joke became a PR lesson: challengers can punch up, leaders can’t — mike64_t · 2026-07-26
- “Cooking AGI” turns into a checkpoint that only learns shallow pattern matching — flowersslop · 2026-07-26