AI audiobooks – 10% more audiobooks this year than all of 2023 | Hacker News Reader

AI audiobooks – 10% more audiobooks this year than all of 2023

bookrank.io

10 pointsbye16 comments

Am I missing where in the article is mentioning AI? I can't find any instances of "AI", "Artificial", or "generated". This seems like a particularly editorialized title.

Fair, I definitely editorialized a bit. It just seemed like an obvious reason to me, but perhaps I am guessing a bit too much at the causes?

A bit is an understatement in this case.

> Otherwise please use the original title, unless it is misleading or linkbait; don't editorialize.

The growth seems entirely in line with the curve. I'm not sure it needs much explaining, let alone guesses.

I don't see how this is linked to AI.

The article doesn't mention AI at all. We know that the number of books being released has exploded in general due to AI, but are those books also being released in audiobook form? Or is this increase due somehow to the use of AI in audiobook recording?

The article does hint at this - with some of the top authors recording hundreds of books - so maybe these folks have used AI to clone their voices and fast track the process?

Personally, I think AI has great potential here, even if only to fill in the gaps. Older, less loved books aren't recorded.

Additionally, there are several narrators that simply grate on my ears, or who my brain has simply learned to tune out, and it would be nice to have an option to switch to a (lower quality) AI version.

I've been enjoying HFY on yt and manga recaps on yt, makes good podcast type content. Also some of the books I listen too are about 6 months behind the audiobook release, so I just use TTS so I can listen while I work/drive.

Audio is great when you multitask, and with the current tv shows lacking in good sci-fi content, books are about the only choice.

How do you manage the TTS on the books?

Youtube is flooded with bad Science Fiction - I assume all literature genres are the same?

From February 2023: "Sci-fi magazine has to halt submissions after receiving too much AI-generated fiction"

https://qz.com/clarkesword-neil-clarke-chatgpt-ai-q-and-a-18...

Yes, 100 times as much, some say

There was a myriad of crappy fiction before LLMs - way more than anyone can read - so… nothing really changed? Sure, LLMs have their own distinct flavors of crappiness, but it doesn’t really matter.

If I got it right, the core problem is that because writing became easier/more accessible, more people got… bolder (for a lack of better word) and started to spam^W publish subpar quality works more, I assume, hoping to make some money out of it.

If so, this seems to be simply a growth-related average quality issue that was surely known for centuries (cf. Mill’s “the general tendency of things … is to render mediocrity”; and if that’s a stretch, then, at the very least, I believe Lem was quite explicit about it coming in his Summa Technologiae)

True, there was always crappy writing, but now an ocean of spamwrite can flow so easily

And that's just a result of growth. There are more humans than ever, communication technologies are improving a lot every decade, writing tools are also improving a lot, and now there are LLMs that can aid with writing. The more accessible it becomes, the more stuff we'll have out there, and most of this stuff is going to be not so great (we'll still have gems, of course). So, it was bound to happen.

The issue you're seeing ($platform spamwrite) is not really an "AI" problem (though, of course, any tech improvement is a sort of an enabler) - its root cause is modern advertising being not aligned with your interests. Incentives there are purely economical (driving sales is the only thing that matters there), and they don't have anything to do with the quality or consumer satisfaction, at least not directly. So, no surprise, when all that matters is consumers' attention (and not the quality of the work), people try to make buck by making the least necessary amount of effort and spamming hard, hoping for an "engagement".

A venal AI assisting a venal, but limited, human..

What we currently call "AI" (LLMs) cannot be venal. It is not conscious and has no agency, so the term is simply not applicable.

It's like calling a rock "murderous" when someone used it to murder another person - understandable, possibly poetic, but completely semantically wrong.

My point is, shifting blame on progress is misdirected and benefits nobody. Fixing issues requires understanding and addressing root causes. As greed (the most fundamental issue, I believe) can be only be possibly solved by achieving post-scarcity (or so I think), I'm suggesting that the next best thing to blame is misaligned incentives and modern economy of "engagement". And not "AI" or "Internet" or anything progress-related.

In essence you are right, a wandering AI - off it's leash will 'run' hither and yon with no guidance - however with a guiding mind it will bark up whatever 'trees' it's master directs..

Reply on news.ycombinator.com