In this video, I break down a story that honestly creeped me out: a grieving widower posts his wife’s final moments on Reddit, and within minutes that raw, confessional post can effectively become AI training data. That’s not hypothetical—Reddit’s 2024 licensing deal with Google (reported at $60 million) is part of a much bigger pattern where Big Tech turns our public—and sometimes painfully personal—words into proprietary intelligence. And the uncomfortable truth is we usually “agree” to this in seconds, with the real consent buried in thousands of words of terms and conditions. I walk through how this has already played out across platforms: Zoom quietly added language in 2023 that allowed AI training on calls and chats (then walked it back after backlash), Amazon’s Alexa recordings have been stored and analyzed since 2014, and there was even a period where contractors manually reviewed snippets for labeling. Tools like Grammarly and Google Docs follow the same general dynamic: by default, your usage can help train models, even if no human is “reading” it line-by-line. My big takeaway is the economics—individually our posts aren’t worth much, but in aggregate they’re incredibly valuable, and we’re not the ones getting paid. I don’t have a tidy conclusion, but I do think we’ll look back and wonder why we gave away so much, so early.