Survey: Nearly Two-Thirds of White-Collar Workers Miss Pre-AI Work Methods
A survey by consulting firm Adaptavist reveals that approximately 65% of white-collar workers miss working without AI, with many citing concerns over reduced creativity and AI misuse. Additionally, nearly half of the respondents reported spending extra time verifying AI-generated content, which contradicts the promise of AI reducing repetitive tasks. This highlights the friction in AI adoption, showing that generative AI can sometimes decrease workplace efficiency due to the "verification tax" and lead to homogenized creative outputs. It also reflects how shifting corporate AI policies—driven by token-based pricing models—are causing frustration and uncertainty among employees. Studies from institutions like the Wharton School of the University of Pennsylvania support these findings, warning that relying solely on AI tools like ChatGPT leads to a lack of diverse ideas. Furthermore, companies that initially encouraged AI usage are now restricting it to control costs as tech providers charge based on token consumption.
## BACKGROUND
Generative AI tools process and generate text using "tokens," which are basic units of text (characters, words, or subwords) that models analyze for semantic relationships. Because AI providers charge businesses based on the number of tokens processed, corporate costs can scale rapidly with high usage, leading to budget caps. Additionally, because LLMs are trained on existing data patterns, their outputs tend to converge on average or similar ideas, potentially reducing creative diversity when overused.