wenbin

Let me guess - the keyword here is "Section 174", just from the title alone :)

Dealing with Section 174 amortization in those first one to three years is a real headache (and your tax bill ends up higher than if it didn’t apply). Once your startup survives that the first few years of doing Section 174, things do get easier... but, sadly, most don't make it that far.

There could be a better system (for me): a read-it-later app with a strict limit—say, 3 or 5 slots. If it's full and you add a new article, the oldest one gets automatically deleted to make room. That way, you always have something to read, but never feel overwhelmed by an endless backlog.

One issue I've had with read-it-later apps is that I end up accumulating far too many articles and never actually reading them. Now my approach is simple: if I see something I want to read, I either read it immediately or never.

Congrats on the launch.

This reminds me of https://www.heavy.ai/ (previously MapD back in 2015/16?)

besides speaking good english, what else do non-technical folks bring? maybe a good family name? /s

Yes, if I need (relatively) accurate answers (with the sources / urls / web pages), I'd use keyword search on Google.

Still have a trust issue with LLM/ChatGPT for facts. Maybe in a couple years my mindset will shift and trust LLM/chatgpt more.

i’ve built a few web apps. the closest to a “vending machine” is https://Transcript.New – dead simple, transactional, zero outages in 14 months, and almost no support emails (maybe 2 in the past year).

If you manage to curate a large enough collection of AI-generated playlists, would you consider putting it on Kaggle? This way, others could use it to train AI models to automatically detect AI-generated music.

We did something similar by sharing a dataset of around 7,000 AI-generated fake podcasts on Kaggle:

https://www.kaggle.com/datasets/listennotes/ai-generated-fak...

It’s great to see more companies adopting fair pricing models.

The first time I encountered this was with Slack—they only charge for active users.

We follow a similar approach with our products : 1) PodcastAPI.com - If no API requests are made within a month, the user pays $0. 2) website Premium Membership- rather than forcing users to pay a monthly fee, we allow them to buy a 2-day pass with one time payment (default option) - listennotes.com/premium

Caveat: customers will demand more. soon, they’ll request for hourly fair pricing - don’t charge me for those hours that i don’t use your service!

Awesome!

The best we’ve achieved is 0 downtime for read operations and less than 1 minute downtime for write ops [1]

Achieving 0 downtime for write ops is super hard!

[1] https://www.listennotes.com/blog/a-practical-way-to-upgrade-...

soon, you’ll see a lot of ai-generated fake podcasts in podcast player apps. Here’s a list of such fake podcasts (several thousands, many are in Apple Podcasts) : https://www.kaggle.com/datasets/listennotes/ai-generated-fak...

Good job! If you need episode search, check out PodcastAPI.com

Cloudflare is truly awesome!

They offer incredibly generous infrastructure components for individuals and small businesses.

If you’re looking to host a podcast with a custom domain name and need significant free storage, you’ll quickly realize there aren’t many (if any) free options—until you discover Cloudflare. With tools like R2 and Pages, they open the door to a world of possibilities.

I’ve even built an open-source podcast CMS/hosting solution using Cloudflare [1]. Thanks to R2, you can host up to 10GB of audio for free! It’s a game-changer.

[1] microfeed.org

AI-generated music can be acceptable in certain contexts, such as white noise for coding, running, or working out, where the primary goal is background ambiance rather than active listening.

However, AI-generated fake podcasts are a different story—they're often frustrating and a waste of people's time.

Unfortunately, platforms like Apple Podcasts and Spotify are increasingly populated with these spammy, AI-created podcasts. For those interested, here's a small subset (4,000+) of such AI-generated fake podcasts available on Kaggle: https://www.kaggle.com/datasets/listennotes/ai-generated-fak....

hetzner is cheap, but cheap often has hidden costs / risks.

aws, azure, and gcp aren’t cheap , but they offer better stability—both technically and operationally.

This is awesome! Is there an example list from Snack.xyz?

People love building lists on the internet.

Browser bookmarks are lists.

Podcasts/RSS feeds are lists.

YouTube has Watch Later / playlists.

Newsfeeds are lists.

Pinterest has a lot of lists of images.

This is definitely not a new issue.

What I’m aiming for is to ensure that the NotebookLM team is aware of the impact and actively considering it. Hopefully, they are already working on tools or mechanisms to address the problem—ideally before their colleagues at YouTube and Google Search come asking for help to fight NotebookLM-generated spams :)

It's certainly easier for the creators of genAI to build detection tools than for outsiders to do so. AI audio detection is a hard problem - https://www.npr.org/2024/04/05/1241446778/deepfake-audio-det...

It’s definitely more than that.

The 1,300+ shows are just the ones recently removed from Listen Notes.

Give it a few days, and I’m sure the number will double, quadruple, and continue to grow. :(

NotebookLM is contributing to fake podcasts across the internet, with over 1,300 and counting:

https://github.com/ListenNotes/ai-generated-fake-podcasts/bl...

Google is taking a different approach this time, moving quickly. While NotebookLM is indeed a remarkable tool for personal productivity and learning, it also opens the door for spammers to mass-produce content that isn't meant for human consumption.

Amidst all the praise for this project, I’d like to offer a different perspective. I hope the NotebookLM team sees this and recognizes the seriousness of the spam issue, which will only grow if left unaddressed. If you know someone on the team, please bring this to their attention - Could you please provide a tool or some plain-English guidelines to help detect audio generated by NotebookLM? Is there a watermark or any other identifiable marker that can be used?

Just recently, a Hacker News post highlighted how nearly all Google image results for "baby peacock" are AI-generated: https://news.ycombinator.com/item?id=41767648

It won't be long before we see a similar trend with low-quality, AI-generated fake podcasts flooding the internet.

Great question.

We only need to provide the sitemap (with custom paths, not publicly available) in a few specific places, like Google Search Console. This means the rules for managing sitemaps are quite manageable. It’s not a perfect setup, but once we configure it, we can usually leave it untouched for a long time.

At Listen Notes, we rely heavily on Cloudflare to manage and protect our services, which cater to both human users and scripts/bots.

One particularly effective strategy we've implemented is using separate subdomains for services designed for different types of traffic, allowing us to apply customized firewall and page rules to each subdomain.

For example:

- www. listennotes.com is dedicated to human users. E.g., https://www.listennotes.com/podcast-realtime/

- feeds. listennotes.com is tailored for bots, providing access to RSS feeds. Eg., https://feeds.listennotes.com/listen/wenbin-fangs-podcast-pl...

- audio. listennotes.com serves both humans and bots, handling audio URL proxies. E.g., https://audio.listennotes.com/e/p/1a0b2d081cae4d6d9889c49651...

This subdomain-based approach enables us to fine-tune security and performance settings for each type of traffic, ensuring optimal service delivery.

If you want something simple, try this one: https://www.microfeed.org/

The argument is that it's fine to generate AI podcasts for personal use, but please refrain from polluting the public space. Statistically, the majority of AI-generated podcasts found on podcast platforms are produced by spammers, scammers, and blackhat SEOs, which are not for human listeners to consume.

I find Notebook LM helpful for generating content for personal projects, like creating my own 'podcasts' on topics I choose.

But would you listen to 'podcasts' generated by others using Notebook LM?

Feel free to subscribe to some of the 'podcasts' listed in this CSV file. Give them a listen and see if they resonate with you.

At Listen Notes, we recently removed over 500 fake podcasts generated by Notebook LM in just the past weekend.

It's disappointing to see scammers and black-hat SEOs already leveraging Notebook LM to mass-produce fake podcasts and distribute them across various platforms.

In the near future, a significant portion of YouTube videos and podcasts will likely be AI-generated (e.g., through tools like Notebook LM).

However, I'm uncertain whether audiences will truly enjoy this AI-generated content. Personally, I prefer content created by humans—it feels more authentic and engaging to me.

It’s crucial for AI tools to include robust detection mechanisms, such as reliable watermarks, to help other platforms identify AI-generated content. Unfortunately, current detection tools for AI-generated audio are still lacking - https://www.npr.org/2024/04/05/1241446778/deepfake-audio-det...

[Edit] We just put together a list of notebooklm generated "podcasts": https://github.com/ListenNotes/notebooklm-generated-fake-pod...

Consider whether you'd enjoy listening to AI-generated podcasts. I believe people might be okay with shows they create themselves, but are less likely to appreciate 'podcasts' ai-generated by others.

If I understand correctly, Automattic and WP Engine generate approximately the same revenue?

The creator of WordPress would undoubtedly be extremely disappointed… they should have achieved much greater success.

Do people actually enjoy this type of AI-generated "podcasts" vs human-produced shows?

As a podcast listener, I lose interest if I can tell the audio is AI-generated...

Still working on ListenNotes.com and PodcastAPI.com —it’s been a 7-year journey so far.

When I was younger, I admired those who could build 12 startups in 12 months. Now, I have more respect for the ones who stick with the same project for years (or even decades) :)

I'd start with a public company that has very few customers.

For example, Palantir