subtech

This is sick!

The initial image is just 1 -- the very first frame of the video.

Used ChatGPT to generate the script lol.

If an international student is on a stem opt extension (24 months) that just started, and does YC, is it possible for them to get a green card before the stem opt expires?

tortoise

Hi HN!

I built a 24/7 Twitch stream where an AI clone of Paul Graham is doing a live AMA with the chat.

You can ask PG any question in the chat, and he'll answer you live.

I built this live stream over the weekend as a fun little hack, hope you enjoy it! :D

Tech stack:

- ffmpeg

- GPT-4

- ElevenLabs / Tortoise TTS

- Wav2Lip

Have fun!

Hi HN!

I built a 24/7 Twitch stream where an AI clone of Paul Graham is doing a live AMA with the chat.

You can ask PG any question in the chat, and he'll answer you live.

I built this live stream over the weekend as a fun little hack, hope you enjoy it! :D

Tech stack:

- ffmpeg

- GPT-4

- ElevenLabs / Tortoise TTS

- Wav2Lip

If you like the stream, I would greatly appreciate any donations to cover server costs: https://www.buymeacoffee.com/thecooltechguy

Have fun!

underrated reply :)

Honestly, thanks for taking the time to give such detailed feedback!

The current model I have isn't as good as I want it to be, and I'm working on a newer one with a different training process as well. This should address some of the shortcomings people have mentioned.

I was really nervous of shipping my v1 with the current model, but thought I might as well share what I have so far with the world, in case someone finds it useful lol.

Yup, working on it!

Autoplay coming!

Thank you so much! With an improved model, things should get a lot better! :D

Playlist + some sort of auto-playing mechanic is coming!

Noted! A lot of people have asked for a playlist support, definitely shipping that soon!

ty for ur kind words! <3

I definitely have to improve the UI, but for now, as you type in a song in the search bar, it should load some auto-complete suggestions that you can choose from.

I had to train my own model :)

Playlists coming very soon!

Hmm I actually use Firefox myself, but tbh, it can be a bit flaky atm due to the sudden surge in traffic. Should become more reliable soon!

Working on some fixes for the duplicate track issue! Sorry about that!

<3

Working on a better music player UI that should also fix this!

I just wanted to quickly THANK EVERYONE for taking the time to check this project out and give your feedback!

I honestly didn't expect this project to get this much traffic -- I really can't express my emotions via text rn lol.

I'm working on an improved AI model that should address a lot of the shortcomings of the current one, along with a lot of other features people have mentioned (playlists, deduplicating results, volume control, better search UX, etc.).

Got a lot of updates coming, time to ship! :D

100% This is definitely worth exploring, and I'm currently trying to figure out the appropriate front-end UI/UX to expose this functionality for users.

Thanks! :)

Let's just say that setting up the backend systems involved a lot of tears & frustration lmfao

Hey thanks for your feedback! :D

Ya Apple Music can have some interesting duplicates in their catalog lol :/

Thanks! <3

The current model does tend to do well with EDM, but got a new model in the works that should hopefully address a lot of the shortcomings of the current one!

I think I can technically use the underlying model to do it, but never really tried it. Interesting...

FAISS for nearest neighbor vector search.