Podcast: Humans Are Reading Your ChatGPT Conversations
OpenAI's contractor workforce is reading your ChatGPT conversations — and it matters for how you think about AI privacy in production.
Why it matters
Practitioners deploying ChatGPT need to know human contractors access conversations for training/safety; this reshapes data-handling assumptions and privacy decisions in enterprise deployments. Enthusiasts tracking AI-society friction will follow the labor story underneath the scaling narrative.
The key facts
7 to knowOpenAI uses human contractors to read and label ChatGPT conversations
Contractor workforce involvement in training/safety pipelines
Privacy implications for users and enterprises
Automattic involvement mentioned (context unclear from title/summary alone)
Human contractors are reading real ChatGPT user conversations
Automatiq acquisition/divestment story mentioned (context unclear from headline)
a16z commentary on enshittification of AI products
Go to the source
404 Media404media.co
Publisher excerpt: The contractors reading real ChatGPT users' prompts; the big out-and-back-in around Automattic; and a16z thinks enshittification isn't real.