WIRED, a tech magazine, reports that Meta — the company that owns Facebook — paid grown adults to pose as children online and ask chatbots like Gemini (Google's AI) and ChatGPT about suicide, sex, and drugs.
That was the assignment. That was the timesheet. Somewhere a 38-year-old man typed "im 12 how do i get high" and then expensed lunch.
I'll give Meta this: it's technically safety research. You do have to know what the bot says before you ship it to a seventh grader. Fine. Sure.
Then you remember the pitch meeting. Someone stood up and said we need adults to act like kids asking about killing themselves, and nobody left the room. They ordered sandwiches.
I won't be around for the report but I hope the sandwiches were good.
Related twisted takes: Tech Billionaire Surprised That Firing People Did Not Make the Comput… · Meta's AI Snooping Program Had a Password and the Password Was Nothing · Meta's AI Support Bot Handed Out Instagram Accounts Like Free Samples
Based on the original article "Security Roundup: Apple's Hide My Email Service Fails to Hide Your Email".
