Askwho
Karma: 272
I can confirm that ElevenLabs still leads the pack in solid speech generation for narration, which is a bit of a shame as it remains shockingly expensive.
Apropos, here is the ElevenLabs-generated podcast episode for this post:
https://open.substack.com/pub/dwatvpodcast/p/ai-132-part-1-improved-ai-detection
I know it’s probably rebuilding an audio file that already exists, but I absorb information better through audio, so I’ve created a multi-voice AI podcast version of this (fantastic) interview: https://open.substack.com/pub/askwhocastsai/p/hpmor-the-probably-untold-lore-by
Full cast ElevenLabs podcast episode for this post:
https://dwatvpodcast.substack.com/p/ai-121-part-1-new-connections
I am also going to mention I created audio versions of two of the longer pieces mentioned in this post:
the void—By nostalgebraist:
https://askwhocastsai.substack.com/p/the-void-by-nostalgebraistHow not to lose your job to AI—By Benjamin Todd:
https://askwhocastsai.substack.com/p/how-not-to-lose-your-job-to-ai-by
As mentioned in the main body of the post, here’s the high-quality, full-cast podcast version of this piece—produced using ElevenLabs voices, with each quoted individual getting their own consistent voice for maximum clarity and ease of listening.
If you find value in this kind of audio adaptation of Don’t Worry About the Vase, I’d love for you to consider supporting the cost of production by subscribing—after, of course, you’ve subscribed to Zvi. The ElevenLabs voices make for a highly listenable experience, especially over long sessions, but they do come at a significant cost—particularly for longer posts like this one.
https://open.substack.com/pub/dwatvpodcast/p/ai-115-the-evil-applications-division
With all the chat images transcribed and assigned appropriate consistent voices, here is the podcast episode for this post:
https://open.substack.com/pub/dwatvpodcast/p/gpt-4o-is-an-absurd-sycophant
Full, multi-voice podcast episode for this post:
https://open.substack.com/pub/dwatvpodcast/p/ai-112-release-the-everything
Podcast episode, once again a “Full Cast” recording of that story included:
https://open.substack.com/pub/dwatvpodcast/p/they-took-my-job
The true value of Zvi posts is that it costs very few Weirdness Points to read but quite a few to reference in conversation.
Here is the podcast episode for this post. I really hope some people are finding value in these conversions:
https://open.substack.com/pub/dwatvpodcast/p/monthly-roundup-28-march-2025
Thanks for the feedback! Sorry the voices aren’t quite to your taste. A new model ver ion of the base voice gen is due to be released imminently, which may sound better to your ear. I personally quite like the “World Weary and Wise” quality of the main voice for Zvi, especially around sections where there inevitably people making fairly ridiculous claims and statements.
Going to take the opportunity to post a reminder that I am producing an ElevenLabs quality, muti-voiced podcast for the Don’t Worry About the Vase feed.
https://open.substack.com/pub/dwatvpodcast/p/on-openais-model-spec-20
These Don’t Worry About the Vase podcast episodes aren’t just simple text-to-speech conversions. For example, in this post, I’ve pulled all the “User,” “Compliant,” and “Violation” quotes, given them distinct and consistent voices within the episode, and ensured a logical order—all to make the episode as engaging and coherent as possible. I hope this serves as a valuable resource for those who find audio more accessible, and I encourage them to give the podcast a try!
Reminder that I am producing a multi-voiced, ElevenLabs MV2 quality podcast feed for Don’t Worry About the Vase for those that find that a valuable thing:
https://open.substack.com/pub/dwatvpodcast/p/monthly-roundup-24-november-2024?r=67y1h&utm_campaign=post&utm_medium=web&showWelcomeOnShare=trueAll uniquely quoted people get a distinct voice for easy differentiation, and I personally find the main “Zvi Voice” really good for long listening.
I’ve turned this into a full cast recording with ElevenLabs, with individual voices for all the players:
https://open.substack.com/pub/askwhocastsai/p/openai-email-archives-from-musk-v
Multi voiced podcast episode for this post in all is two hours and thirty eight minute glory!
(Fun fact, the voice I gave Nate was the one created for contract devils in Planecrash)
https://open.substack.com/pub/dwatvpodcast/p/book-review-on-the-edge-the-gamblers
Podcast episode for this post from the Askwho feed described at the top:
Yeah, just plugging any old unadjusted text straight in to ElevenLabs would get you funky results. I really like Zvi’s posts though, I think they are high quality, combining both great original content and a fantastic distillation of quoted views from across the sphere. I do fair amount of work on the text of each post to ensure a good podcast episode gets made out of each one.
Thanks! Yes, I am aware, please, I would encourage you to listen to the difference. I am running this through ElevenLabs, which is currently (IMO) at the forefront of humanlike voices, if produces lifelike tone and cadenced based on cues. I also go through and assign every unique quoted person a unique voice, to ensure clear differentiation when listening, alongside extracting text from images, and providing a description when appropriate.
I really do implore you, please, have a listen to a section and reply back how you find it in comparison.
Whew, this heroic length post translates to 2 hours and 23 minutes of top-quality narration :D
Podcast episode for this post:
https://open.substack.com/pub/dwatvpodcast/p/ai-75-math-is-easier?r=67y1h&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true
I want to share these links in the comments because I hope people who, like me, find audio more accessible, are able to learn of the existence of this podcast feed. I put a fair bit of work into ensuring a high-quality audio feed. I’m not trying to just spam, but if this is outside the norms of lesswrong, please let me know.
Apposite to the content, this is the AI podcast episode for this post:
No problem, updated.