Reads like it was aided by AI (LLM) in the writing. No description of how the analysis was done. So I have to imagine that jobs are extremely simply counted, n

Reads like it was aided by AI (LLM) in the writing. No description of how the analysis was done. So I have to imagine that jobs are extremely simply counted, not deduped. So a job listed in-office or in-city, but also listed as remote, would be 2 jobs and this would result in 50% remote jobs available. Whereas this is actually 100% remote. So I imagine the 69% is an undercount, if going by my assumption of simplistic counting.