lblume

13 ms·

121.
by lblume 1y ago
Yes. In Germany the formula n(n+1)/2 is actually called the Gaussian sum formula, or even the "small Gauss". [0] [0] https://de.wikipedia.org/wiki/Gaußsche_Summenformel
122.
by lblume 1y ago
What would a sane government even look like at this point?
123.
by lblume 1y ago
> stop wasting time and money on things like that What do you mean? The AMA? > listen to the community Huh? Isn't that exactly what they are doing with this?
124.
by lblume 1y ago
I didn't mean typical chatbot output, these are luckily still fairly recognizable due to stylistic preferences learned during fine-tuning. I mean actual base model output. Take a SOTA base model and give it the first two paragraphs of
125.
by lblume 1y ago
Sure, maybe solving NLP is too great a claim to make. It is still not at all ordinary that beforehand we could not solve referential questions algorithmically, that we could not extract information from plain text into custom schemas of str
126.
by lblume 1y ago
Imagine telling a person from five years ago that the programs that would basically solve NLP, perform better than experts at many tasks and are hard not to anthropomorphize accidentally are actually "trivial". Good luck with that
127.
by lblume 1y ago
The linked page claims that most identified crawlers are related to scraping for training data of LLMs, which seems likely.
128.
by lblume 1y ago
Given that current LLMs do not consistently output total garbage, and can be used as judges in a fairly efficient way, I highly doubt this could even in theory have any impact on the capabilities of future models. Once (a) models are capabl
129.
by lblume 1y ago
Exactly. It looks fascinating, but it looks like there is no further explanation or exploration of the patterns, as if something was cut off.
130.
by lblume 1y ago
It is easier to design nets that only catch big cods and don't also catch too much garbage.
131.
by lblume 1y ago
Technology and memes spread faster than genes. Mosquitoes adapting that fast is not an issue that serious. Additionally, if hundreds of thousands of Americans died of malaria each year, we would have found much more effective cures (conside
132.
by lblume 1y ago
But how often do you actually need that when writing a reactive app? In most cases I have worked with you want the button to persist and load some external data that updates a reactive state.
133.
by lblume 1y ago
Really? So far I have never been disappointed with Cargo, or the Rust toolchain in general. For my work it has been a frictionless experience.
134.
by lblume 1y ago
You can also calculate the XOR-accumulation of all values between 1 and n in O(1) using a lookup table like this: [n, 1, n+1, 0][n%4]
135.
by lblume 1y ago
I have uploaded entire books to the latest Gemini and had the model reliably accurately answer specific questions requiring knowledge of multiple chapters.
136.
by lblume 1y ago
But don't these new costs create a direct incentive to cooperate?
137.
by lblume 1y ago
Thanks a lot for your work!
138.
by lblume 1y ago
Wow, this is incredibly useful. I was just thinking "why can't C-r just work like fzf", but this is a much better approach of course.
139.
by lblume 1y ago
It is hard to write macros in LaTeX.
140.
by lblume 1y ago
Especially in combination with file watching. Your script writes to the JSON file and the entire document and everything that depends on it updates automatically, often in less than a second.
141.
by lblume 1y ago
The Typst compiler is completely open-source. I prefer my local copy of the Typst compiler and CLI to whatever LaTeX provides right now already, and there seems to be a still growing community that could keep the project going even after a
142.
by lblume 1y ago
A slight bias in favor of the status quo might be acceptable or even desired. However current LLMs strongly favor traditional languages and are unable to comprehend even modern language features not part of their base training set. Consider
143.
by lblume 1y ago
Anything is preferable to Microsoft Word.
144.
by lblume 1y ago
I have so far not been able to get a major LLM to generate fully functional Typst code, no matter how much context I try to put into it. The models do not seem to currently understand Typst's concept of modes (code, markup and math) an
145.
by lblume 1y ago
You seem to use the word "good" very liberally.
146.
by lblume 1y ago
The star appears to have been cut off from the link, leaving an article about the corresponding radio source, Sagittarius A. This (official Wikipedia) short link leads to the desired article: https://w.wiki/5A7e
147.
by lblume 1y ago
Moreover, the bias could be arbitrarily subtle. This is something that really worries me about black box AI systems.
148.
by lblume 1y ago
The difference is that V8 is sandboxed.
149.
by lblume 1y ago
Was Jekyll ever that popular though?
150.
by lblume 1y ago
No, in fact it is available since iOS 4 (2010). You just connect the account and turn on the notes synchronisation (right below calendar synchronisation).

More ›