Outside of a web scraper, how sure are we that this poisons reddits actual data being sold to ai companies? It seems trivial for them to have an original comment field in the database that’s invisible to users or just use backed up data. Or even an anonymized copy of all all original comments not linked to any account that is solely for AI training.