gary_0

56 ms·

841.
by gary_0 4y ago
The issue of the Web's lack of distinction between document viewing features and being an application platform (with access to system resources and private data) is decades old. Unfortunately, the question has been academic for decades
842.
by gary_0 4y ago
Techniques like NeRF allow you to take a bunch of photos of a real 3D scene and then generate images/video of the scene from arbitrary viewpoints, where NeRF will infer the 3D structure using machine learning. So what you're seein
843.
by gary_0 4y ago
Anyone know how many input images were used to train the network for the house scene? I don't see that information anywhere.
844.
by gary_0 4y ago
Opus goes down to 5ms if you customize it, and there's no point in going any lower for Internet streaming because 5ms of raw audio data is only 480 bytes at 16-bit 48khz (plus lots of UDP packet overhead). For latency-sensitive local s
845.
by gary_0 4y ago
If I can continue the nitpickery, plenty of computer algorithms are deterministic but also pseudorandom--including Markov chains. Only when the randomness is outside your control are things truly non-deterministic. A simulation with many
846.
by gary_0 4y ago
> the upside of infinity dollars It would be better to use the market cap of the currently dominant company in the sector to compute the upside (unless the startup is doing something completely novel, like autonomous robotics or nuclear
847.
Creating a minimal ELF-64 file using Clang and Rust(github.com)
1 points by gary_0 4y ago|0 comments
848.
by gary_0 4y ago
Subtle data corruption due to bit-flips isn't necessarily solved by backups. You might not notice until the damage is already done, and having old good data somewhere would be no help at all.
849.
by gary_0 4y ago
Sure, and he also has long-term plans to build a Tesla autonomous android, and a SpaceX Mars colony. I wish him the best of luck, but I'm not going to put money on any of those plans succeeding any time soon.
850.
by gary_0 4y ago
Elon/Twitter stories get clicks in part because of the "how's he going to get out of this one!?" factor. As the article points out, Twitter's financial numbers and upcoming liabilities don't look good. This is
851.
Intel Battlemage, Celestial GPUs Are Booked in at TSMC(tomshardware.com)
1 points by gary_0 4y ago|0 comments
852.
by gary_0 4y ago
I had a feeling my number was out of date, but wow!
853.
by gary_0 4y ago
Humans have top-notch visual/spatial reasoning systems and we still bang our shins and elbows on stuff. We shouldn't be removing proximity sensors from cars, we should be installing them on people!
854.
by gary_0 4y ago
To the end user, it won't matter much either way, because most open source software already compiles and runs just fine on RISC-V; there are plans for projects like Chromium and Android to fully support RISC-V. RISC-V is on a trajector
855.
by gary_0 4y ago
In addition to what others have said, C and C++ rely on a textual macro preprocessor and text-based #include, which can lead to a lot of annoying code-breakage when refactoring. Automated tools can't parse C/C++ unless they'r
856.
by gary_0 4y ago
Agreed. There are a hundred different kinds of information processing that go into a human-like mind, and we've kinda-sorta built one piece. And there are a lot of pieces that it would neither be sane nor useful to build (eg. internali
857.
by gary_0 4y ago
I think that includes laptops, where macOS is very popular? For gaming, at least, Steam's stats show 96% Windows.
858.
by gary_0 4y ago
And it takes ~20 years to train a new brain so it can coherently answer questions about a wide variety of topics. Even worse, you can't even copy-paste it once you're done!
859.
by gary_0 4y ago
Microsoft has 95% of the desktop OS market, they can abuse their users with impunity until that number starts to go down.
860.
by gary_0 4y ago
That's how the brain evolved, after all: take incoming input, process according to previous patterns, output a reaction to peripheral nervous system that will enhance survival. The evolutionary purpose of a cortex is to store a statist
861.
by gary_0 4y ago
Most of the effort that went into making humanity's data structured, searchable, semantic, and linkable on the open Web has been violently unwound in the past decade or so. Now, huge amounts of data are stashed behind some social netwo
862.
by gary_0 4y ago
IBM has more employees than Google. It might only make a quarter of Google's revenue, but that's still a lot of money. That's a big zombie.
863.
by gary_0 4y ago
> Especially because Google has exactly one revenue source. Google is an advertising company that just happens to have a lot of engineers on its payroll.
864.
by gary_0 4y ago
Google is too big to die anytime soon, but they're already dead to me. The company that seemed so cool in 2004 is long gone. They're just another Oracle or IBM or Microsoft now.
865.
by gary_0 4y ago
In the end I think LLM's will not be used on their own much, and instead more as glue for bringing together a whole bunch of other technology and allowing it to be coordinated with plain English.
866.
by gary_0 4y ago
This amusing thought of yours is more plausible than I'd like it to be. Yikes.
867.
by gary_0 4y ago
Yeah, this is a common problem I run into with blog.whatever, forum.whatever, or wiki.whatever pages. They never bother customizing Wordpress, MediaWiki, etc so that you can navigate to the main site. Fer chrissakes, the Web was more usable
868.
by gary_0 4y ago
(2020). I don't see a full date on the article (we can probably blame SEO for that), but archive.org says it's from April 14, 2020.
869.
by gary_0 4y ago
Mac is a lot of money to spend if you care about good Linux support, and Raspberry Pi makes for a very weak workstation (if you can even get one for a reasonable price). The demand for a developer-oriented ARM workstation is there, but the
870.
by gary_0 4y ago
It's fun how many events I'm living through that are on "worst X in history" lists.

More ›