Lockal

19 ms·

211.
The developer of ComfyUI/StableDiffusion3 4B has resigned from Stability AI(old.reddit.com)
1 points by Lockal 2y ago|0 comments
212.
by Lockal 2y ago
That's a big jump in generalization that bruteforcing 4 colors in 9x9 grids with 8000 programs has anything near to what sentient human can do. Back in the days similar generalization was used for Deep Blue chess computer. Computer won
213.
by Lockal 2y ago
And even if platform starts to make blurry transitions for ads, adjust volume/brightness, or even overlay ad text over existing video like a watermark, client can just download the same video multiple times and remove unique frames&#x2
214.
by Lockal 2y ago
Not sure what we can conclude from this graph. Why it is not normalized? https://pubmed.ncbi.nlm.nih.gov/?term=illness - try any common word and you will see that it grows just because of number of papers. https:/&#x2
215.
Kostya/benchmarks: Some benchmarks of different languages(github.com)
1 points by Lockal 2y ago|0 comments
216.
by Lockal 2y ago
Note that libdivide may sound like an improvement, however there are better solutions today: 1) full-blown compiler/engines like LLVM or GCC build close to optimal shift/mul-based code (without overhead for simple constants like &
217.
Squares in Squares(kingbird.myphotos.cc)
2 points by Lockal 2y ago|0 comments
218.
by Lockal 2y ago
Sounds like another item in this list https://elonmusk.today
219.
by Lockal 2y ago
"Influence operations": Oh no! We were removed! What should we do? Oh, wait, there is "Sign up" button
220.
by Lockal 2y ago
You are saying as if automatic detection of misinformation and brigades was ever possible before LLMs, or it is possible now. Only now, when people started to massively notice things like https://news.ycombinator.com/item?id
221.
by Lockal 2y ago
Even assuming that bitcoins are indeed were directly mined, it's ridiculous how asymmetrical the modern world is about some things: "I (personally) issued counterfeit bills and was able to purchase real estate with them" - th
222.
by Lockal 2y ago
I think you massively underestimate the complexity of pytorch. Even if we exclude all GPUs except for AMD, and exclude clang (required for AOT engine), pytorch depends on almost every ROCm library. And inside it depends on original Triton l
223.
by Lockal 2y ago
Maybe due to 2x slower matmul (well known bottleneck), as one of multiple factors? https://github.com/tinygrad/tinygrad/blob/v0.9.0/extra/gemm/... There are multiple bounties just for it in ht
224.
by Lockal 2y ago
Tinygrad targets consumer hardware (to be precise, only Radeon 7900XTX and nothing else[1]), while ROCm does not actually provide good support for such hardware. For example, last release of hipBLASLt-6.1.1 library has deep integration with
225.
by Lockal 2y ago
"editorialise = express opinions rather than just report the news" It is a anti-clickbait practice to remove "you" or "we" (implying reader) from titles on HN, not editorialization.
226.
by Lockal 2y ago
"It is safe", they said, but now they see nothing more than https://en.wikipedia.org/wiki/Fetal_alcohol_spectrum_disorde... variant. Surely pharmaceutical lobby will try to solve the situation with even more
227.
by Lockal 2y ago
It sounds funny, but no real evidence provided that it actually works. Here is more real example: https://youtu.be/aLx2q-UnH6M?t=1621 - user injected "SIMA Balls" into result, but other than that, there were many
228.
by Lockal 2y ago
Hmm, what a comparison, in one case a tyrant, pursued by the doctrine of "Russian world" ( https://en.wikipedia.org/wiki/Russian_world ) kills hundreds of thousands of people for the sake of annexing territory.
229.
by Lockal 2y ago
The article is not useless as many might think. It is a pity that it is not as well thought out as it appears. For example, the suggestion to remove system applications will almost immediately lead to a bootloop. I, for example, have a hard
230.
by Lockal 2y ago
Better "best" - https://hn.algolia.com/
231.
by Lockal 2y ago
Yes, I think you are right. I just forgot that "As an AI language model" is a signature from ChatGPT too. Sometimes it reveals itself with phrases like "is against OpenAI's use case policy" (which I don't see i
232.
by Lockal 2y ago
Example articles generated with LLM: https://archive.ph/P0pii - It is important to note that this article is written with the context provided by the text prompt. It highlights the cynical tone towards the US government, NA
233.
by Lockal 2y ago
Related: https://www.empatheticfiring.com/
234.
by Lockal 2y ago
This is incredibly stupid, but I think we've reached the stage where it's time to add a new rule to the https://news.ycombinator.com/newsguidelines.html page: "links must be working (i.e., show content that m
235.
by Lockal 2y ago
> They are providing this tool for users to run Where? From what I see in https://deepmind.google/technologies/synthid/#:~:text=Where%... , it is not a tool, it is an "undescribed algorithm". Otherwis
236.
by Lockal 2y ago
Ah, that flexible "COST license" - "contact our sales team" license
237.
by Lockal 2y ago
"Oracle’s move seems like a straightforward decision to ensure it is using the most permissive underlying IaC tool without having to worry about downstream license complications, no more and no less." Hm, with that logic they coul
238.
by Lockal 2y ago
... with a proprietary algorithm, the details of which we will not say a word about
239.
by Lockal 2y ago
> may consume several hours of your time Several hours? More like few lives, the stack is just too big to deal within a single life :) And the rabbit hole goes even deeper with external tools like https://alive2.llvm.org/
240.
by Lockal 2y ago
Tone of Ulrich Drepper is remarkable there, that was definitely... a time

More ›