A new benchmark can test how much LLMs become sycophants, and found that GPT-4o was the most sycophantic of the models tested.

Are you imagining things, or do artificial intelligence (AI) chatbots seem too eager to agree with you? Whether it’s telling ...

Three recent incidents are reminders that generative AI tools remain troublesome and unreliable. IT buyer beware.

GPT-4o is ChatGPT’s default, flagship model. It’s a jack-of-all-trades, but a master of none, and should be used for everyday ...

Steven Adler breaks down recent erratic behavior from leading chatbots. It has been an odd few weeks for generative AI ...

Gopnik makes no prescriptive claims in her analysis, but when we think of AI in this way, it becomes evident that in seeking ...

A striking graph, which went viral on Reddit last week, shows the percentage of internet users visiting OpenAI's blockbuster ...

It comes days after the chatbot went off the rails with talk of “white genocide” in South Africa. Read more at ...

OpenAI recently rolled back an update that made Chatgpt overly sycophantic, which worsened this problem by excessively ...

Sam Altman envisions that ChatGPT could remember a user's entire life journey in future as young people use it as a life ...