Mmastodon TechnologyAI first seen 2 h ago, last 2 h ago, peak #3
Awkward mistakes now seen as proof of human writing
Original: 1996: little awkward mistakes in online posts show that you're not fluent in a foreign language. 2026: little awkward mi
A sharp observation circulating online notes a complete reversal in how small language errors are read. In 1996, awkward mistakes in a foreign language online signalled a non-fluent writer; by 2026, similar imperfections suggest the author actually knows the language, rather than relying on machine translation or AI tools. The quip captures growing suspicion that polished text is machine-generated, with imperfect, idiosyncratic phrasing becoming a marker of genuine human authorship and multilingual skill.
Why now: The spread of AI and machine translation has flipped perceptions, making human-sounding errors a badge of authenticity online.
David Megginsonmachine translationartificial intelligence
Evidence
- 1996: little awkward mistakes in online posts show that you're not fluent in a foreign language. 2026: little awkward mistakes in online posts show that you actually know a foreign language (rather than relying on machine translation). # language # AI # bilingual · david_megginson@mstdn.ca · 2
API: https://socialmediatrends-api.osmike.com/v1/trends/1727497