mohamedkoubaa

Please contact me or share your contact information, I am working on something exactly like this.

I know people who got into car accidents and 30 years later they were fine

GPU programs as they are currently authored. It's because there isn't a runtime or DSL at the right conceptual level that is still practically useful against modern hardware and APIs. This is what I wanted to tackle with goldy.

Goldy has the high level structure of the shader graph and exchanges with the outside world (host memory, surfaces, etc) which is turned into a backend specific graph IR (different backends have different rules for command list retention and other properties). But my library also embeds Slang and uses Slang IR to inject and introspect on the user written shaders. I can then at runtime turn it into CUDA graphs (slang can turn any shader into a CUDA kernel) but with the added benefit that the runtime can manipulate the kernels and graph as more information is available. For example if a host upload that usually is scheduled between kernels isn't done on a specific graph submission, the graph partitioner can fuse kernels and remove a fence.

I plan to take advantage of the compiler and runtime introspection to implement some interesting features like shader coroutines without sacrificing GPU residency.

I do something related in my gpu library. After a few frames if the push constants don't change I compile the shaders in the background with them defined out by the preprocessor to reduce the size of the shader program (kind of like a branch predictor). I also store the entire pipeline in a graph data structure that I partition into segments that let me fuse and split kernels (though I hadn't implemented those optimizations yet). In my mind one issue with these GPU accelerated programs is that there isnt a runtime with the right level of information about the overall program to do compiler style optimizations, especially for complex programs.

(The library is called goldy, and until I spend some time on it the readme and docs are sadly LLM generated)

It doesn't hold knowledge it has the ability to seive through a lossy latent representation of knowledge.

I love it. Entropic prompt injection

The question might be less about what replaced the world reserve currency and more about how the global economy might function or not function without a world reserve currency

At this point Microsoft should just open source it for science.

The browser as an implementation of the standard is not differentiable. But a universal app engine with sandboxing untrusted apps, high performance, and platform neutrality that may or may not retain a DOM or render HTML is very browser-like and under explored.

On the other hand, I don't know if a slightly more customizable browser is something that's really worth building from scratch.

Plans for android?

People who produce good software don't feel the need to show off how modern and agentic they are.

I get flamed here whenever I say it's a skill issue, but it absolutely is. Not everyone has what it takes to be a successful CTO but that's the role you play if have 20 agents working on something you designed. I see you got downvoted too.

We need a push notification: "Living room TV wants to use its microphone" and for it to be locked out if you don't allow it.

>there is no sign of the AI slowing down or being unable to maintain it

This means you designed the system well and made sure the agent implemented it correctly. Even at Big Company the big shot architects don't review every line of code, and nobody seemed to mind that

My point is that it isn't that RN "removes one of the biggest annoyances of developing for the App Store", it's that a specific technical feature of RN does that, and it's not only available to RN

This seems to be a weaker claim than "RN is needed to do this", but practically speaking it might amount to that.

Pardon my ignorance but can't an iOS app host a WASM runtime without being a browser? And if so it doesn't need to be an interpreted language necessarily.

Are you saying a native app can't make an http request and update the code they JIT?

I'm in the same field and I find GPT to be better at understanding physics conceptually but Claude is better at writing numerical code. I use cursor so many of my sessions start in GPT and switch to Claude.

I would vote for "A bit of a pickle"

I came across an interesting article once that claimed that Buddha was reconstituted into a mythical St. George

They are being removed from hundreds of municipalities, what do you mean?

Associating the term policing with legitimacy is not something every person defaults to

"Works on my machine" accelerationism

Some of these are just office buildings with multiple companies. Are we really calling a rented space with servers a data center now?

This is some bull

I think you should focus on the fundamentals, this kind of feedback is inevitable and I'm sure you anticipated it anyways.

This'll be resolved by a selection process. Code that gets quadratically more complex will at some point fail to advance even with access to hundreds of genies.

Eh.. Using page rank in 2026 is like using a bow and arrow in 1918

The race to the bottom continues