A gentle reminder to my fellow Long Covid advocates using AI for research or drafting posts: please proofread medical AI output very closely.
These models were trained on data heavily shaped by the dominant narratives of institutional PR and academic marketing, meaning they will sometimes confidently explain away systemic failures and even emphasize and congratulate some of the very "experts" who have arguably contributed to six years of inaction and abandonment of patients with pandemic-related conditions.
If we don't independently audit the output, we risk accidentally validating and amplifying the exact misinformation we are fighting so hard to change.💕