> LLMs are innately “converging” — their underlying dynamics push them towards objective reality — citing the example of Elon Musk’s repeated comeuppances at the hands of fact-checks by his own AI chatbot, Grok.
Yes, but people have confirmation bias, and will ignore contradictions in a chatbot, and public figures, in favor of what confirms their biases.
For example: Who tweeted the following in 2019?
The United States has spent EIGHT TRILLION DOLLARS fighting and policing in the Middle East.
Thousands of our Great Soldiers have died or been badly wounded. Millions of people have died on the other side. GOING INTO THE MIDDLE EAST IS THE WORST DECISION EVER MADE..
https://archive.is/20260328080158/https://www.ft.com/content...
> LLMs are innately “converging” — their underlying dynamics push them towards objective reality — citing the example of Elon Musk’s repeated comeuppances at the hands of fact-checks by his own AI chatbot, Grok.
Yes, but people have confirmation bias, and will ignore contradictions in a chatbot, and public figures, in favor of what confirms their biases.
For example: Who tweeted the following in 2019?