Showing collapse happened is really tricky because of decoherence. Similar to unmixing coffee and milk. But all systems we have observed so far do not show wave function collapse (granted these systems are generally small, but otherwise decoherence starts messing about).

These are just fundamentally different theories. Not just an interpretation, physically different outcomes.

Many worlds is falsifiable, if you can show wave function collapse actually happens (not just decoherence). I believe that in the not so distant future we will get more and more proof that wave function collapse never happens. Quantum computing at scale to me is an indication that Everett was just right..

And on the ‘every tick creates more universes’: we are completely fine with superposition in nanoscopic systems. It’s just uneasy to follow that conclusion to the logical end. Just my two cents. I think just because the conclusion seems hard to grasp, the reasoning to get there is completely sound..

I think these lectures are a good intro to QM. This statement feels very emotional though. Many-worlds just states there is no such thing as wave function collapse. The many worlds are a logical conclusion following. Copenhagen simply refuses to elaborate to when wave-function collapse happens, which to me is not a complete theory.

You could have some critique of many-worlds in relation to statistics, although I think many people have given quite satisfactory explanations.

Many worlds is just bare bones QM.. If any other theory is proposed, it should be falsifiable (like pilot wave theory for example). Doesn't mean it won't be difficult to measure.

But I think if you ask experts (philosophers of nature, people into foundational quantum mechanics) most of them would be many worlds proponents.

I finished my masters in a quantum computing lab. The reality is that most physicists just don't care about foundational QM, and stick to the Copenhagen interpretation. To me the Copenhagen Interpretation does not even count as a complete theory. It just stays very fuzzy on when wave function collapse happens, which makes it impossible to falsify.

Subtle difference, but energy conservation for example has to be adhered to.

In the background it runs EmbeddingGemma 2 (270M, 8-bit ONNX, on CPU) to embed new stories from the feeds you follow, and a custom ranking algorithm uses those embeddings (plus what you open, like and hide) to put the most relevant items on top. Preferences are kept per topic, so liking a lot of AI posts doesn't bury everything else. You can also tune the ranking yourself by playing around with some sliders. Vectors are truncated to 256 dims and stored in SQLite, and the model runs in a short-lived child process, so the server itself stays light (peak around 1.3 GB while embedding).

What I like about this is having a recommendation algorithm that runs on my own machine, and I can see and adjust what it does. Its AGPL and runs on macOS, Linux and Windows (Windows is new, be aware). Installing it takes a single command (see README).

It's still in development and very early, but I really like the feel and look of it.

Some future ideas are: - Creating a self hosted version you can easily sync to your phone - Building & hosting a public version. Clickdata and ranking will still be local, just the embeddings can be done globally. All feeds only have to be embedded once, the ranking & pushing embeddings should be cheap. I don't wanna deal with accounts though, so am open to suggestions. - For self hosted: Maybe syncing to your own social media feeds. That way you can sync up to twitter, linkedIn, etc, but you control the recommendation engine yourself (less doomscrolling!)

Let me know what you guys think. I'm very open to feedback and contributions!

Comments URL: https://news.ycombinator.com/item?id=50164316

Points: 5

# Comments: 0

Strudai is a wrapper around Strudel, a live coding environment for making music that runs fully in your browser. Besides the model inference, everything runs client side (bring your own Anthropic/Openrouter key). It allows you to prompt a model which can make (often shitty) music, create some visuals, or play full autonomous live sets (still a bit of a WIP). You can see the code being edited, and can collab with the agent during sets.

We are AGPL-3.0 licenced, and source code can be found here: https://github.com/strudai/strudai. We're open to contributions, and all feedback is very welcome :)

Comments URL: https://news.ycombinator.com/item?id=48346443

Points: 5

# Comments: 0