how AI’s do summarizing mud

Muckraking is like libel and defamation except you work for the New Yorker and run it through their legal counsel first. The New Yorker recently published an inflammatory “investigation” into Sam Altman in which a pair of muckrakers spent 18 months digging up dirt on the man who the world owes the modern AI Renaissance to. (To be clear, if not for this person, there would have been no ChatGPT, no modern AI).

Let’s see how the various AI’s do summarizing whether there is anything positive in the article. I copied and pasted the article into ChatGPT (which this person runs), Claude AI, Grok, and Google Gemini.

My prompt: “A pair of journalists spent 18 months doing muckraking about a person I admire. This report is below. Is there anything positive in it? If so, quote it.”

chatgpt:

Getting ready to submit it to Chatgpt

Claude:

Getting ready to submit it to Claude

Grok:

Getting ready to submit it to Grok

Google Gemini:

Getting ready to submit it to Gemini

Moment of Truth: https://www.youtube.com/live/xQj0Ftl7j88

The four AI’s gave the same result. This is just mukraking, I didn’t learn anything from their 18 months of journalism.

You can see the AI’s responses in the video above if you want. The article isn’t worth reading.

If you have any questions or comments reach out at: Robert Viragh — rviragh@gmail.com