Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
filterfiber
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
91.
▲
by
filterfiber
3y ago
Does anyone know roughly based on the specs how this roughly compares to the pi family for performance?
92.
▲
by
filterfiber
3y ago
They also have a decent advantage for LLMs because of their memory bandwidth to system memory vs GPU's with limited VRAM limited over PCIE to the system memory.
93.
▲
by
filterfiber
3y ago
Is this an old build or the modern engine? According to steam charts they have 70.1k players in-game right now. I'm intrigued that the 11th highest current player count on steam has native linux support and open sourced their engine? T
94.
▲
by
filterfiber
3y ago
Not quite a city building but I think possibly comparable - factorio you can at least get to endgame on pretty low end hardware with decent fps. It even runs on the switch. I say comparable because of the number of things moving around on b
95.
▲
by
filterfiber
3y ago
It is 12bit but the bottom 0-0.1v and 3.2v-3.3v just read as the min-max. Honestly for many projects needing analog it might be "good enough", but it leaves a lot to be desired, and like the parent said, the moment you're loo
96.
▲
by
filterfiber
3y ago
That makes sense thank you!
97.
▲
by
filterfiber
3y ago
How do you feel about the esp32's? I know there's a few variants of the 32 with different capabilities, but how do you feel about using that as a general purpose MCU including projects that might not need wifi/bt? I'm so
98.
▲
by
filterfiber
3y ago
> you have an untrusted binary and you want to analyze it to determine if it's malicious, or hook its encryption function so you can decrypt the traffic and see what it's sending out of your machine. Program debuggers do this i
99.
▲
by
filterfiber
3y ago
Does anyone have any experience for how the generalized LLM's like llama/gpt4 compare to much smaller, purpose built summarizers like some of the FLAN variants?
100.
▲
by
filterfiber
3y ago
It's likely from the RLHF Basically on A/B tests, humans tend to prefer more saturated, "punchy" images. Which is also why iPhones tend to do the same thing. For artificial images people also seem to prefer stylized &quo
101.
▲
by
filterfiber
3y ago
> If you actually care about privacy rather than want to pay lip service to some marketing, I disagree that it's lip service Brave has a ton of engine level privacy patches https://github.com/brave/brave-core&#x
102.
▲
by
filterfiber
3y ago
> It's like installing land mines in a residential neighborhood and then blaming the residents for stepping on one. It's not even reasonable if you warn them. No this is seriously a DMZ. I cannot stress this enough. Seriously a
103.
▲
by
filterfiber
3y ago
The anticheat isn't defective here. It's like running across the DMZ. Your intentions don't matter - ALL anticheat will ban you on sight. You MUST collaborate with the developers to not get banned. This is widely known. It is
104.
▲
by
filterfiber
3y ago
> The anti-lag system is not a cheat. The anti-cheat system thinks it is. By definition, it's a false positive. Yes but this is not a false positive you fix. This is a DMZ - you do not cross it no matter what your intentions are. Th
105.
▲
by
filterfiber
3y ago
No it's definitely not. Reading yes (antivirus). Writing no, that's rare and generally purposeful. I have no idea what processes you have that do that unless they're specialty tools like reshade, autohotkey, etc. Programs do
106.
▲
by
filterfiber
3y ago
So yes it is a false positive, but it's not the type VAC is concerned about. What AMD did was like crossing the DMZ. It doesn't matter who you are or what your intentions are. It's not possible for normal applications to do t
107.
▲
by
filterfiber
3y ago
> graphics pipelines compile shaders though? From my understanding the newer api's have built in methods to negotiate these optimizations. And the gpu driver is free to do whatever it likes after the api calls. So the game tells the
108.
▲
by
filterfiber
3y ago
To my understanding the graphics drivers modify the rendering pipeline after the draw calls are made, they don't actually hijack code within the game and modify it's logic. Developers of games add api calls to their game engines f
109.
▲
by
filterfiber
3y ago
From the game's perspective this was a cheat. The anti-cheats job is to make sure nobody tresspasses with the game logic. They modified where methods returned to so they could run their own code before returning to the game. VAC cannot
110.
▲
by
filterfiber
3y ago
> Which is to say that their anticheat system is defective because the method it uses has false positives -- as exemplified by this case, which is not an attempt to cheat. This is not a false positive. What AMD did is identical to what c
111.
▲
by
filterfiber
3y ago
> the only (unfortunately) novel thing here is doing this without the express consent and cooperation of the game's developer. Nvidia did it "properly" and worked with valve to develop it in-engine. I could definitely be
112.
▲
by
filterfiber
3y ago
They certainly will if they haven't already. Last I heard they were waiting for amd to fix their code before they unban people. This is the third false-VAC since the cs2 beta. First was a unprotected yaw command (no commands should get
113.
▲
by
filterfiber
3y ago
This is extremely wild to me. AMD literally made an "internal cheat" for cs2. This is a widely known guaranteed ban and likely among most people (if you modify the game engine then you get banned). Even with weak anticheat like VA
114.
▲
by
filterfiber
3y ago
> It's important to recognize that predicting text is not merely about guessing the next letter or word, but rather a complex set of probabilities grounded in language and context. When we look at language, we might see intricate re
115.
▲
by
filterfiber
3y ago
The average person sadly just hears the marketed "artificial intelligence" and doesn't grasp that it simply predicts text. It's really good at predicting text we like, but that's all it does. It shouldn't be su
116.
▲
by
filterfiber
3y ago
> Does Bard clearly warn to never rely on it for facts? I know OpenAI says "ChatGPT may give you inaccurate information" at the start of each session. I know I shouldn't be, but I'm surprised the disclosure is even ne
117.
▲
by
filterfiber
3y ago
> Does anyone suck at that? Does android even have native OCR? Last I checked everything required an OCR app of varying quality (including windows/linux). On ios/macos you can literally just click on a picture and select the te
118.
▲
by
filterfiber
3y ago
Not OP, In my experience apple's ML on iphones is seamless. Tap and hold on your dog in a picture and it'll cut out the background, your photos are all sorted automatically including by person (and I think by pet). OCR is seamless
119.
▲
by
filterfiber
3y ago
Do you use wayland, multiple monitors, and/or play games or is it just for ML/AI?
120.
▲
by
filterfiber
3y ago
> There's transparent evidence that transformer-based LLMs existed before ChatGPT or Llama. Transformers were first published about in "Attention is all you need" in 2017 by google https://arxiv.org/abs
More ›