I don't have access to the DeepMind demo, but from the video it looks like it takes the idea up a notch.

(I don't know the exact lineage of these ideas, but a general observation is that it's a shame that it's the norm for blog posts / indie demos to not get cited.)

[1] https://news.ycombinator.com/item?id=43798757

[2] https://madebyoll.in/posts/world_emulation_via_dnn/demo/

Comments URL: https://news.ycombinator.com/item?id=46668200

Points: 2

# Comments: 1

"Tell HN: Viral Hit Made by AI, 10M listens on Spotify last few days" [1]

[1] https://news.ycombinator.com/item?id=46600681

> I couldn't resist putting a face to this gem.

After saying it is AI:

> How people was touched by this version of « Papoutai » by stromae made by AI, I was also touched as y’all, as an independent Artist I wanted to put my emotions and my soul on this masterpiece

https://www.youtube.com/watch?v=bQ8GbwQV5zE

The author has stated it's AI in their YouTube video descriptions [1] [2], e.g.

> Disclaimer: I do not own the rights to the original song's lyrics or melody. This is an Al-generated cover intended for entertainment and homage. The original composition, lyrics, and music are the sole property of the original artists, writers, and copyright holders.

It has gone viral in last few days, due to TikTok and YouTube Shorts.

10,000,000 listens on Spotify - https://open.spotify.com/track/0HI7kOxEfnSMD2Z27BJnJN

There's various other YouTube videos of it with millions of views.

I heard of it myself through YouTube shorts, and after a few listens suspected it was AI.

[1] https://www.youtube.com/watch?v=OWHpljeJv34

[2] https://www.youtube.com/watch?v=bQ8GbwQV5zE

Comments URL: https://news.ycombinator.com/item?id=46600681

Points: 4

# Comments: 5

    bulk_send(
        generate_expiry_email(user) 
        for user in db.getUsers() 
        if is_expired(user, date.now())
    )

Yep this kind of thing can happen. I found and reported incorrect gradients for Apple's Metal-backed tensorflow conv2d in 2021 [1].

(Pretty sure I've seen incorrect gradients with another Pytorch backend, but that was a few years ago and I don't seem to have raised an issue to refer to... )

One might think this class of errors would be caught by a test suite. Autodiff can be tested quite comprehensively against numerical differentiation [2]. (Although this example is from a much simpler lib than Pytorch, so I could be missing something.)

[1] https://github.com/apple/tensorflow_macos/issues/230

[2] https://github.com/sradc/SmallPebble/blob/2cd915c4ba72bf2d92...

- Convert the whole codebase into a string

- Paste it into Gemini

- Ask a question

People seem to be very taken with "agentic" approaches were the model selects a few files to look at, but I've found it very effective and convenient just to give the model the whole codebase, and then have a conversation with it, get it to output code, modify a file, etc.

Nice synergy here, the lineage is: Karpathy's nano-GPT -> Keller Jordan's modded-nanoGPT (a speedrun of training nanoGPT) -> NanoChat

modded-nanoGPT [1] is a great project, well worth checking out, it's all about massively speeding up the training of a small GPT model.

Notably it uses the author's Muon optimizer [2], rather than AdamW, (for the linear layers).

[1] https://github.com/KellerJordan/modded-nanogpt

[2] https://kellerjordan.github.io/posts/muon/

[1] https://giscus.app/

> Each update gets a timestamp (physical or logical).

> When two devices write to the same field, the update with the latest timestamp wins.

Please also have a robust synced undo feature, so it's easy to undo the thing you don't want that gets written. Apps that sync often seem to be stingy about how much "undos" they store/sync (if any).

So for $3000, that's 3000 hours, or 125 days, (if just wastefully leave them on all the time, instead of turning them on when needed).

Say you wanted to play around for a couple of hours, that's like.. $3.

(That's assuming there's no bonus for joining / free tier, too.)