AI Resilience Report forWord Processors and Typists

Today, many typing jobs are partly aided by AI. Software at work automatically checks spelling, grammar and punctuation. In fact, research finds that automated writing tools give quick feedback and even reduce workload for people checking text ^[1].

Likewise, voice-to-text systems on computers and phones are now very good – studies of modern speech-recognition show error rates around 8–10% ^[2]. That means most spoken words can be typed by AI as accurately as a human transcriber. However, some tasks remain manual.

For example, adjusting page layout or choosing formatting styles usually still needs a person’s touch (word processors have templates, but not smart AI to do it all). Physically collating printouts or sorting mail often stays hands-on. Overall, the number of dedicated typist jobs has fallen a lot.

The BLS reports about 37,000 word-processors/typists in 2023 ^[3], down sharply from roughly 305,000 in 2014 ^[4]. This drop reflects how technology now handles many routine clerical tasks that people used to do.

AI tools for typing work are already available, which can speed up adoption. Built-in dictation and grammar checkers exist in phones and word apps, so companies can use them without building new tech ^{[2] [1]}. Since these workers earn about $22–$23 per hour on average ^[3], an AI system that does similar work could be seen as cost-saving (especially if it cuts errors or works faster).

On the social side, younger users and students are already comfortable with AI writing tools ^[1], suggesting future office workers will be too. Still, human oversight is important. Tasks needing judgment or taste – like deciding on tone or fine-tuning a layout – are harder to automate.

Companies will likely use AI to handle heavy routine parts (making offices more efficient), while skilled humans focus on proofreading, new writing or tasks that need creativity and careful decision. This way, humans and AI work together.