Tom Simonite
simonite.bsky.social
Tom Simonite
@simonite.bsky.social
Tech companies editor @washingtonpost.com // [email protected]
And that ChatGPT began responses with variations of “yes” or “correct” almost 10 times as often as it started with “no” or “wrong.”

More on how ChatGPT behaves in this story on how a tool pitched mostly for productivity has been embraced by users for much more
www.washingtonpost.com/technology/2...
We analyzed 47,000 ChatGPT conversations. Here’s what people really use it for.
What do people ask the popular chatbot? We analyzed thousands of chats to identify common topics discussed by users and patterns in ChatGPT’s responses.
www.washingtonpost.com
November 13, 2025 at 7:07 PM
Sora can mimic Netflix shows, movies, TikTok videos and Twitch streams, suggesting it was trained on versions of the originals. Netflix said it didn't provide data to OpenAI. Twitch said it doesn't have a partnership w/ OpenAI.
September 19, 2025 at 4:32 PM