|
I think you're right here. I'm forgetting about "entertainment". This will never replace my visits to HN because it's not efficient. Unless it uses neurolink to beam the information into my mind, that's always going to be the case. It fails in that way. However, let's consider a podcast. You get 5 people that are entertaining, knowledgeable about technology, well spoken, and have differing opinions, and get them to spend 30 minutes talking about the top HN posts. It's slow, they would most likely not even discuss my favourite posts, but yet, I would consider listening. Is it possible to replicate that with AI? Could AI digest the content and all of the comments, and turn it into an entertaining and educational discussion and debate between a few different AI voices? I think it's possible, and it changes the idea from being an inefficient method of summarizing content, to a form of entertainment. That could be the right direction to go with this kind of project. However, even if AI perfectly replicates a great podcast, how would I feel listening to it? I think this question applies to most AI content. How important is the human process in the content we consume? Do people only care about the end result, or do they want read a book written by a person, view photos taken by a person, listen to songs performed by a person, and listen to a podcast by real people? |
I was thinking more in terms of the way you can get morning news briefs on a nest hub. The AI summarising the top posts on HN would have value to me as something to listen to in the morning not because I expect a really interesting debate, but mostly just to get a quick update on what the buzz is on HN on a given day so I can decide if I want to pick up the thread from any of those topics when my day starts at the desk.