Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Google Illuminate: Books and papers turned into audio

illuminate.google.com

Post reply on HN
Print view View original

181–190 of 251 posts

Re: Google Illuminate: Books and papers turned into audio

#181

Post by hliyan »

freefaler wrote: post #3

Great idea. I wonder how long until we'd see a lot of "autogenerated" podcasts with syndicated advertising inside spamming the podcast space. Like with robovoiced videos on YT reading some scraped content.

I'm conflicted about this. On one hand, it makes content more accessible to a larger audience. On the other hand, it leverages copyrighted material without crediting or compensating creators, potentially puts those same creators out of work, and finally, reduces the likelihood of more such (human) creators arising in the future. My worry is that a few generations hence, human beings will forget many skills like this, and if model collapse occurs due to LLMs ingesting their own data over successive iterations, future generations will be in for a difficult time. Reminiscent of Asimov's "The Feeling of Power".

Re: Google Illuminate: Books and papers turned into audio

#182

Post by netdevnet »

smeej wrote: post #42
ertgbnm wrote: post #38

Earlier quoted context omitted.

Depends on the what you are trying to get out of a podcast. Most of the podcasts I listen to are because I want to learn something new in an entertaining format. I'm not listening to develop parasocial relationships with the hosts, so removing that element could be a good thing for me. Of course if you listen to podcasts because you like the parasocial aspect or the celebrity interviews, then yeah... Not really a poi…

I don't know that "parasocial relationships" are the primary reason people like having real hosts. I have a huge list of things I've managed to change in my life because I heard some other real person talking about how they were possible. Listening to these people over time and realizing there's nothing about them that's so special that it makes things possible for them that aren't possible for me gets me off my butt…

> I don't know that "parasocial relationships" are the primary reason people like having real hosts

But it is likely one of the main. Me telling you that something is possible doesn't necessarily mean that it is real but you chose to believe it. Whether the source is human is not necessarily relevant. After all humans can and do lie all the time

Re: Google Illuminate: Books and papers turned into audio

#183

Post by mavhc »

hliyan wrote: post #181
freefaler wrote: post #3

Great idea. I wonder how long until we'd see a lot of "autogenerated" podcasts with syndicated advertising inside spamming the podcast space. Like with robovoiced videos on YT reading some scraped content.

I'm conflicted about this. On one hand, it makes content more accessible to a larger audience. On the other hand, it leverages copyrighted material without crediting or compensating creators, potentially puts those same creators out of work, and finally, reduces the likelihood of more such (human) creators arising in the future. My worry is that a few generations hence, human beings will forget many skills like this,…

If they forget they can find an AI generated youtube tutorial to learn it

Re: Google Illuminate: Books and papers turned into audio

#184

Post by the_duke »

fallinditch wrote: post #142
mmsc wrote: post #135

Earlier quoted context omitted.

This is a bit meta for me. A year ago a website was posted on here HN which allowed you to visit a random website with an /ideas page. For some reason it would always land me on the same website, which outlined something close to this. The idea was something like an RSS feed that would summarize all the entries in the feed for the day/weekin the form of a podcast. I wonder if that was inspiration for Wondercraft.

Uncanny, maybe ;-) It should be easy to ingest an RSS feed into your personal RAG system's vector database. Then set up speech to text and text to speech for your PrivateGPT so you can then ask it to create a podcast just for you

There is no truly great open source TTS system available, as far as I know.

Most of them are rather scrappy.

Re: Google Illuminate: Books and papers turned into audio

#186

Post by hakonslie »

fallinditch wrote: post #43
freefaler wrote: post #3

Great idea. I wonder how long until we'd see a lot of "autogenerated" podcasts with syndicated advertising inside spamming the podcast space. Like with robovoiced videos on YT reading some scraped content.

Wondercraft have been offering this service for a while, and produce some of their own auto-generated podcasts including the Hacker News Recap which does an excellent job of summarizing the most engaged posts on HN. https://www.wondercraft.ai/our-podcasts

I tried listening to the Hacker News Recap a while back, but it was extremely boring and not helpful at all for me.

Re: Google Illuminate: Books and papers turned into audio

#187

Post by rmbyrro »

vanishingbee wrote: post #71
dlisboa wrote: post #12

One problem I see with this is legitimizing LLM-extracted content as canon. The realistic human speech masks the fact that the LLM might be hallucinating or highlighting the wrong parts of a book/paper as important.

Happens in the very first example: [Attention is All You Need - 1:07] > Voice A: How did the "Attention is All You Need" paper address this sequential processing bottleneck of RNNs? > Voice B: So, instead of going step-by-step like RNNs, they introduced a model called the Transformer - hence the title. What title? The paper is entitled "Attention is All You Need". People are fooling themselves. These are stochastic p…

In a sense they are parrots. But the comparison misses cases where LLMs are good and parrots are useless.

Re: Google Illuminate: Books and papers turned into audio

#188

Post by authorfly »

vanishingbee wrote: post #71
dlisboa wrote: post #12

One problem I see with this is legitimizing LLM-extracted content as canon. The realistic human speech masks the fact that the LLM might be hallucinating or highlighting the wrong parts of a book/paper as important.

Happens in the very first example: [Attention is All You Need - 1:07] > Voice A: How did the "Attention is All You Need" paper address this sequential processing bottleneck of RNNs? > Voice B: So, instead of going step-by-step like RNNs, they introduced a model called the Transformer - hence the title. What title? The paper is entitled "Attention is All You Need". People are fooling themselves. These are stochastic p…

Agreed. Another example in the first minute of the "Attention is all you need" one.

"[Transformers .. replaced...] ...the suspects from the time.. recurrent networks, convolution, GRUs".

GRU has no place being mentioned here. It's hallucinated in effect, though, not wrong. Just a misdirecting piece of information not in the original source.

GRU gives a Ben Kenobi vibe: it died out about when this paper was published.

But it's also kind of misinforming the listener to state this. GRUs are a subtype of recurrent networks. It's a small thing, but no actual professor would mention GRUs here I think. It's not relevant (GRUs are not mentioned in the paper itself) and mentioning RNNs and GRUs is a bit like saying "Yes, uses both Ice and Frozen Water"

So while the conversational style gives me podcast-keep-my-attention vibes.. I feel a uncanny valley fear. Yes each small weird decision is not going to rock my world. But it's slightly distorting the importance. Yes a human could list GRUs just the same, and probably, most professors would mistake or others.

But it just feels like this is professing to be the next, all-there thing. I don't see how you can do that and launch this while knowing it produces content like that. At least with humans, you can learn from 5 humans and take the overall picture - if only one mentions GRU, you move on. If there's one AI source, or AI sources that all tend to make the same mistake (e.g. continuing to list an inappropriate item to ensure conversational style), that's very different.

I don't like it.

Re: Google Illuminate: Books and papers turned into audio

#189

Post by Tepix »

keyle wrote: post #154

I listen to 5 mins of this and all I can feel is sadness and how cringe it is. Please do not replace humanity with a faint imitation of what makes use human, actual spontaneity. If you produce AI content, don't emulate small talk and quirky side jabs. It's pathetic. This is just more hot garbage on top of a pile of junk. I imagine a brighter future where we can choose to turn that off and remove it from search, like…

If AI-generated speech is robot-like, dull and monotonous, it will be boring. I think we need human-like speech to make it interesting to listen to. What's your solution to this problem?

OTOH, i think the AI generated stuff should be clearly marked as such so there is no pretending.

Post reply on HN