Unverified50% confidenceOpinionExact time
Some commenters believe that when Dijkstra was writing, machines' ability to understand natural language was essentially zero, whereas today large models can handle ambiguity and context to a considerable degree, so EWD667 criticizes "treating natural language as a substitute," and the more realistic current approach is "natural language as the interface, formal verification as the safeguard."
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/23/2026
First Seen
Valid until: 1/21/2027
Sources
Related Entities
Related Claims
UnverifiedWhat Dijkstra opposed was using natural language itself as a programming language, not describing requirements in natural language.75% similarUnverifiedThe core argument of EWD667 is that natural language is vague, ambiguous and context-dependent, whereas programs require precise, unambiguous, mechanically executable instructions, so treating natural language as a programming language is a misunderstanding of the nature of programming.75% similarUnverifiedThe researchers acknowledge that the model still falls far short of classical language models trained on trillions of tokens on text tasks.74% similarUnverifiedThe current mainstream academic consensus is that existing large language models do not possess consciousness or subjective experience.72% similarUnverifiedThe output of large language models inherently appears plausible (fluency bias), meaning that even when the reasoning chain is wrong, the generated natural language remains grammatically correct and natural in tone, rendering traditional monitoring approaches based on output format or exception capture ineffective.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1091261API
curl https://kongchang.com/api/v1/knowledge/claims/1091261MCP
get_claim(id=1091261)