cubefox

26 ms·

751.
by cubefox 7mo ago
> Companies shouldn't wait to solve issues like this - they should be proactively helping their most vulnerable users. I think they should help their median users and empower their power users, and they should absolutely throw a few
752.
by cubefox 7mo ago
Do people complain about being scammed with Windows or macOS? Apparently not. So they probably also don't complain about Android. The security seems more an excuse to become more closed. Like iOS.
753.
by cubefox 7mo ago
But in order to be compressed, don't we have to load the image into memory first, uncompressed? I don't quite see how this could result in reduced memory usage.
754.
by cubefox 7mo ago
Why would they benefit from hardware compression?
755.
by cubefox 7mo ago
Unfortunately the article mentions nowhere why a GPU would ever need to compress (rather than decompress) images. What's the application for that? In the beginning the article mentions formats for computer game textures, but I'm
756.
by cubefox 7mo ago
Real numbers mostly appear in calculus (e.g. the chain rule in gradient descent/backpropagation), but "discrete calculus" is then used as an approximation of infinitesimal calculus. It uses "finite differences" rath
757.
by cubefox 7mo ago
That's was a great read and I agree with the author. Though to be honest, I don't particularly like the type of Amiga pixel art in the article. That is, pixel art with relatively high resolution and relatively high color depth. Ev
758.
by cubefox 7mo ago
I don't think any PS2 games used texture streaming. They used LOD, which is different. (As I argued below, the asset streaming that Crash Bandicoot used was even simpler than LODs because it relied on a fixed camera path.)
759.
by cubefox 7mo ago
That game had a fixed camera on rails though. Theoretically, they could have used a pre-rendered video of the level and just adjust the playback speed. There are some racing games which actually did that. So the asset streaming could be lin
760.
by cubefox 7mo ago
I wonder whether there could be a way to circumvent this with procedural textures, like the ones we saw in the .kkrieger demo. Alternatively, perhaps the texture streaming could be made to work with PlayStation or Saturn, though I assume th
761.
by cubefox 7mo ago
The original "Lumen in the Land of Nanite" demo is nearly six years old actually (May 2020).
762.
ARC-AGI solver via Plücker geometry – 316 tasks solved with zero learning(github.com)
3 points by cubefox 7mo ago|0 comments
763.
by cubefox 7mo ago
No, they are only using ray traced global illumination, which Unreal Engine already had several years prior (Lumen). They are not second place either, because several other engines also had it before id Tech. > They ripped our Carmacks t
764.
by cubefox 7mo ago
Unfortunately nowadays id Software doesn't seem to be at the cutting edge of engine technology anymore. Most interesting new developments now come from Unreal Engine as far as I can tell. Like virtual geometry (Nanite) or efficient ray
765.
by cubefox 7mo ago
Yes, id invented it, but I think they published one slightly earlier game which also had texture streaming. The technique (virtual textures) would not become ubiquitous in most engines until the PS4 era though.
766.
by cubefox 7mo ago
The same guy, James Lambert, also implemented texture streaming (which would not be invented until two console generations later) in an N64 demo. The textures look uncharacteristically high res: https://youtube.com/watch?v=S
767.
The Terrarium(lesswrong.com)
2 points by cubefox 7mo ago|0 comments
768.
by cubefox 7mo ago
Funny, downthread there is actually a report of someone experiencing manipulation of weather prediction markets: https://news.ycombinator.com/item?id=47535355
769.
by cubefox 7mo ago
Being good at predicting is a broad skill which is applicable widely, while insider information only works for the specific thing the information is about. Far from every market will have insiders with special information. E.g. on who will
770.
by cubefox 7mo ago
Or you are simply better at predicting the future than most other people.
771.
by cubefox 7mo ago
I think they don't actually lack them, or lack only a small fraction (their brains are ≈99% like a normal human brain), such that if they were an AI model, they could be fairly trivially upgraded with vision capability.
772.
by cubefox 7mo ago
Well, it would be AGI if you could connect a camera to it to solve it, similar to how blind people would be able to solve it if you restored their eyesight. But if the lack of vision is a fundamental limitation of their architecture, then i
773.
by cubefox 7mo ago
It's not obvious at all. And I would say pretty much impossible without using machine learning. Even for ARC-AGI-1 there is no GOFAI program that scores high.
774.
by cubefox 7mo ago
This is supposed to test for AGI, not ASI. ARC-AGI (later labelled "1") was supposed to detect AGI with a test that is easy for humans, not top humans.
775.
by cubefox 7mo ago
It's client side scanning.
776.
by cubefox 7mo ago
No because the base model from which the distilled or quantized models are derived is larger.
777.
by cubefox 7mo ago
Some yes, others no. Distillation and quantization can't be used to make new base models since they require a preexisting one.
778.
by cubefox 7mo ago
> What are the most importsnt breakthroughs from the past two or three years for intelligence? The most important one in that timeframe was clearly reasoning/RLVR (reinforcement learning with verifiable rewards), which was pioneere
779.
by cubefox 7mo ago
> But if they read your paper enough that they invited you to a talk, that probably means they were far enough along to independently inventing it That's more than a stretch. They likely invited them because someone thought the abst
780.
by cubefox 7mo ago
> Can you do basic math on the number of support tickets Valve is handling to answer that question for yourself? Can you? Do you even know that number?

More ›