Speech to text processing to improve such that by end of 2025 it will be an essentially solved problem.

Created by JoshuaZ on 2015-02-09; known on 2026-01-01

• JoshuaZ estimated 55% on 2015-02-09
• JoshuaZ said “http://lesswrong.com/r/discussion/lw/lor/discussion_of_concrete_neartomiddle_term_trends/ is context” on 2015-02-09
• EloiseRosen estimated 80% on 2015-02-09
• Houshalter estimated 95% on 2015-02-09
• quicklystarfish estimated 99% on 2015-02-16
• ThetaAurigae estimated 90% on 2015-02-16
• Pablo estimated 85% on 2015-02-19
• two2thehead estimated 95% and said “Siri. Google Now. Cortana. Alexa from Amazon. Now Facebook. Eye ball to bug ratio is pretty high.http://www.wsj.com/articles/facebook-buys-voice-recognition-startup-1420496634” on 2016-05-02
• themusicgod1 said “Also important to note: without context it’s only going to get so good. With context it is probably already solved. The problem is the context.” on 2016-05-02
• two2thehead said “Excellent point themusicgod1. :-)” on 2016-05-05