20k

17 ms·

271.
by 20k 2y ago
Every test I've ever tried with an LLM to get it to generate code has produced a complete unworkable mess. I have no idea what people are generating with them, but its always far less work to write something that I understand, rather t
272.
by 20k 2y ago
There's no scientific evidence that adults can learn absolute pitch with anywhere near the accuracy of people who've had it from a young age, despite a lot of testing. The article we're talking about is a negative result with
273.
by 20k 2y ago
>size of error reduced by 42.7% (from 2.62 to 1.50 semitones) for the trained timbre, which generalized partially to an untrained timbre. Overall, results provide more convincing evidence for the learnability of AP judgment in adulthood
274.
by 20k 2y ago
tl;dr they were able to get an AI to copy a directory and run a program with extensive prompting, by bruteforcing running everything in the directory. It could query a liveness port by running or generating netstat (which is cool!) to check
275.
by 20k 2y ago
Knowing how to do it through the CLI adds nothing to your understanding of the concepts though. Its just whether or not you have to remember an opaque CLI command, or you get a UI to click
276.
by 20k 2y ago
Honestly, 99% of the pain of git is simply because people use it through the CLI. If you use tortoisegit or a visual tool, you don't need to worry about any of this because its self explanatory, and it becomes trivial to use Learning g
277.
by 20k 2y ago
Long covid, and CFS, are almost certainly the same disease. The person you're replying to is massively overselling the efficacy of treatment though, most people with this kind of fatigue tend to improve to a lower level of functioning
278.
by 20k 2y ago
Aka people have been consistently calling out the AI hype as being excessive for decades, despite a weird push by the marketing segments of the programming community to declare everything as being AGI. The current technology is better and h
279.
by 20k 2y ago
The thing is, that's not true at all. AI is great for some tasks, and poor for other tasks. That's the reason to break it down like this, because people are trying to explain where AI will and won't revolutionise things, inst
280.
by 20k 2y ago
Eh it performed a 1:1 conversion of ARM NEON to wasm SIMD, which with the greatest will in the world is pretty trivial work. Its something that ML is good at, because its the same problem area as "translate this from english to french&
281.
by 20k 2y ago
So what it misses is that the ADM formalism is a subset of GR, in that it can only simulate a subset of the full equations for GR. It also gives a lot of reasoning there that is quite faulty with respect to answer the question of if its str
282.
by 20k 2y ago
Its hard to overstate just how wrong chatgpt was about everything that I asked it - from the broad explanations, to the details of it. If you ask it "why does the universe expand" (which is a basic question), it'll be wrong I
283.
by 20k 2y ago
Nearly everything I've ever asked chatgpt about astrophysics has been critically wrong in some fashion or other. The issue is is that all its training data is polluted by popsci Eg I just asked chatgpt about the metric expansion of spa
284.
by 20k 2y ago
Its hard to explain exactly what's accurate here when you're looking at the precise shapes - part of the problem is that this rendering is extraordinarily complex to do in real time, and it relies on the size of the object (tris
285.
by 20k 2y ago
What this thread is really missing is a simulation. I can't promise its the enterprise (as the performance constraints are crippling), but here's a bunch of cubes instead: https://www.youtube.com/watch?v=iTw0pJvTkG
286.
by 20k 2y ago
Its worth noting that adults can learn something called pseudo absolute pitch, which sounds similar to what you're describing. The difference (which you're also describing) is that it requires reinforcement, and is a bit less accu
287.
by 20k 2y ago
So, I've got perfect pitch, and its not quite that black and white. For context, I can do all the regular perfect pitch superpowers (except for door rattles, which don't really rattle with a single tone), and have had it from a yo
288.
by 20k 2y ago
I've been able to port NR to GPU's which with sufficiently powerful hardware can run simulations at about ~30fps with raytracing, to simulate binary black hole collisions. You need something around a top end consumer gpu at the mo
289.
by 20k 2y ago
The preview shows that you have a static black hole selected This shape of the shadow is also wrong for kerr though, this is what kerr looks like: https://i.imgur.com/3cS1fNI.png
290.
by 20k 2y ago
As far as I can tell, the black hole's you're generating don't look especially correct in the preview: they should have a circular shadow like this https://i.imgur.com/zeShgrx.jpeg
291.
by 20k 2y ago
I was having more of a think about this, one of the issues with LCG's is that they're linear, which means that even relatively basic cryptographic analysis can defeat them. The issue here particularly is that, say you have a crypt
292.
by 20k 2y ago
Man that one time pad generation is...... not good, I would bet that its very possible to recover the seed state for the different generators, LCGs are trivial to reverse engineer. It looks like the PRNG's are seeded by the system RNG
293.
by 20k 2y ago
Nvidia have actually been putting a fair amount of work into their OpenCL drivers in the past few years and have a pretty decent implementation, I have a feeling that opencl is more widely used in the embedded space and they were kicked int
294.
by 20k 2y ago
I feel like this article misses why people avoid TCP like the plague, and why people use UDP for many applications 1. Routers do all kinds of terrible things with TCP, causing high latency, and poor performance. Routers do not do this to ne
295.
by 20k 3y ago
Its not that they're supporting buggy code, they just downgraded the quality of their implementation significantly. They made the compiler a lot worse when they swapped to rocm https://github.com/RadeonOpenCompute/