WeltzeitgeistGreat video. Watch it!
(This is Prof. Ada Palmer @adapalmer)
Matthieu Urvoy@wackJackle Thanks for sharing this! Her take on the successive tech revolutions about information is quite interesting.
<my 2cts>
But I find it difficult to consider the output of LLMs/Gen AI as "information". I'd rather word it as "data".
For it to be considered as "information", I would expect these technologies to provide "new facts", which they can't (or hallucinate, in which case they are false information).
Therefore I don't see how this serves the information history.
</my 2cts>
SamuelJohnson@birozularutti @wackJackle If you read up on distinctions between data, information, knowledge and wisdom you'll find no requirement for information to be true, accurate or novel.
Its character is informing, that is communicating something meaningful. Misinformation and disinformation are forms of information often even more amenable to communication ("a lie gets halfway round the world before the truth gets its boots on").
@samueljohnson @wackJackle I need to read a bit more about that. The definitions for data and information I had in mind were close to these: https://bloomfire.com/blog/data-vs-information/
According to these definitions, I still think that GenAI/LLM output does not qualify as information. At best, it mimics information provided by others.
But I agree on disinformation being a form of information.
