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

NotebookLM's automatically generated podcasts are surprisingly effective

simonwillison.net

Post reply on HN
Print view View original

221–230 of 511 posts

Re: NotebookLM's automatically generated podcasts are surprisingly effective

#221

Post by _emacsomancer_ »

sirsinsalot wrote: post #131
robinsonb5 wrote: post #106

Earlier quoted context omitted.

The thing is, we have been here before. Think back to the mid-1980s and the first time everyone got their hands on a Casio or Yamaha keyboard with auto-accompaniment. It was a huge amount of fun to play with, just pressing a few buttons, playing a few notes and feeling like you were producing a "real" pop song. Meanwhile, any actual musicians were to be found crying in the corner of the room, not because a new tool h…

And yet Clint Eastwood by the Gorillaz was a Casio demo track. It isn't so black and white. https://youtube.com/shorts/Wn0NtSNeQEQ

Similarly, Under Mi Sleng Teng[1], but here too it required human musical talent.

[1]: https://en.wikipedia.org/wiki/Sleng_Teng

Re: NotebookLM's automatically generated podcasts are surprisingly effective

#222

Post by throwuxiytayq »

littlestymaar wrote: post #153
richardw wrote: post #45

Earlier quoted context omitted.

I think and hope that you're wrong. There's always been cheese, and there's a lot of it now. But there is still a market for top-notch insight. For example, Perun. This guy delivers an hourlong presentation on (mostly) the Ukraine-Russia war and its pure quality. Insights, humour, excellent delivery, from what seems to be a military-focused economist/analyst/consultant. We're a while away from some bot taking this ki…

Pleased to see Perun being mentioned on HN.

It's always funny when I find out that various people I respect follow Perun uploads closely.

Re: NotebookLM's automatically generated podcasts are surprisingly effective

#223

Post by fny »

richardw wrote: post #45
timoth3y wrote: post #21

NotebookLM's is incredibly good at generating the affect and structure of a quality podcast. This is in-line with all art, music, and video created by LMM at the moment. They are imitating a structure and affect, the quality of the content is largely irrelevant. I think the interesting thing is that most people don't really care, and AI is not to blame for that. Most books published today have the affect of a book, b…

I think and hope that you're wrong. There's always been cheese, and there's a lot of it now. But there is still a market for top-notch insight. For example, Perun. This guy delivers an hourlong presentation on (mostly) the Ukraine-Russia war and its pure quality. Insights, humour, excellent delivery, from what seems to be a military-focused economist/analyst/consultant. We're a while away from some bot taking this ki…

Try Lawfare as a better LLM hurdle. The depth and expertise and at times physical experience required for their discussions seems far out of reach.

I suspect LLMs are not sophisticated enough as a paradigm to get there.

Re: NotebookLM's automatically generated podcasts are surprisingly effective

#224

Post by ionwake »

stevage wrote: post #75

Jesus it's good. I gave it some of my travel blogs, and wow. I mean, there are flaws, particularly in the shallowness of the analysis, but it's at least as good as some time-poor podcast hosts would do.

TBH Im wondering is there anyway to increase the depth or approach by prompting a model for it? Will that be in a future release or hybrid product? ( The audio tech is seemless 100% perfect ) its the quality of the content that needs work now, is there no way to plug this into another LLM ?

Re: NotebookLM's automatically generated podcasts are surprisingly effective

#226

Post by DiscourseFan »

mrits wrote: post #197
DiscourseFan wrote: post #190

Ok, so, this is my impression from shoving philosophy texts into it. For things that already have a large body of scholarship, and have a set of fairly solidified interpretations, it is very good at giving summaries. But for works that still remain enigmatic and difficult to interpret, it fails to produce anything new or interesting. It seems to be a more complex version of ChatGPT, but it has the same underlying pro…

I thought it goes without saying that these types of things depend on existing data. It seems useful to learn about something on your walk or commute but not take your research to the next level.

I'm just saying that, just like other models, it appears at first to have a great deal of use value but in actuality it only works for small edge cases and on things that you can find easily yourself without the model.

Re: NotebookLM's automatically generated podcasts are surprisingly effective

#227

Post by olavgg »

This is really awesome, I just added my startup website as a source, which is a mess of data engineering content written a little bit by myself and mostly by chatgpt 3.5 one year ago. What I find really impressive, it reads the big SVG I have on the landing page, and create a story about a real world use-case scenario.

The result: https://intellistream.ai/static/intellistream_podcast2.ogg

Re: NotebookLM's automatically generated podcasts are surprisingly effective

#228

Post by refurb »

lynx23 wrote: post #52
richardw wrote: post #45

Earlier quoted context omitted.

I think and hope that you're wrong. There's always been cheese, and there's a lot of it now. But there is still a market for top-notch insight. For example, Perun. This guy delivers an hourlong presentation on (mostly) the Ukraine-Russia war and its pure quality. Insights, humour, excellent delivery, from what seems to be a military-focused economist/analyst/consultant. We're a while away from some bot taking this ki…

Seriously, hardcore history? I dont even remember where I heard from him, but I think it was a Lex podcast. So I checked out hardcore history and was mightily disappointed. To my ears, he is rambling 3 hours about a topic, more or less unstructured and very long-winded, so that I basically remember nothing after having finished the podcast. I tried several times again, because I wanted it to be good. But no, not the…

Yes! I’m a huge history buff (read hundreds of books) and was so excited when someone told me about Hardcore History.

I tried a few episodes. I really tried. I couldn’t do it. It reminded me of my uncle would tell a 5 min story in half an hour.

The dramatic filler, breathless story telling was too much for me. If anything it would put me to sleep.

I’ve found a few history podcasts that I think go into a lot more depth and I learn a lot more from.

Re: NotebookLM's automatically generated podcasts are surprisingly effective

#230

Post by karaterobot »

Let's say the use case is that you want to get a light, conversational summary of some dense, technical articles while you're out for a walk. Even if you thought this service was awesome on day one, if you used this every day for a month, would you hate it by the end or not? It's neat, but I can imagine it becoming repetitive quickly, and the seams starting to show after the initial impression wears off.
Post reply on HN