Yes. Suppose you ask me what the sqrt(4) is and I tell you 2. Accurate and correct, right?
Does it matter if I answer every question with either 1 or 2 and flip a coin each time to decide which?
Deterministic means that if it is accurate/correct once, it will continue to be in future runs (unless the correct answer changes; a stopped clock is deterministic).
I think the analogy breaks down here. The elided bit "time indicator" implied at the end makes that statement is false. A stopped clock is not a deterministic time indicator.
If the correct answer changes, a (correct and accurate) deterministic model either gets new input and changes the answer accordingly, or is not correct to begin with.
Determinism is unrelated to correctness. Deterministic means the output depends only on the state you consider to be relevant, and not other factors. A stopped clock is deterministic: no matter what you do, it gives you the same output. A working, accurate clock is deterministic if you consider the current time to be a relevant piece of state, but not if you don't. Consider how "deterministic builds" need to avoid timestamping their build products, because determinism in that context is assumed to mean that you can run it at a different time and get the same result.
LLMs can be deterministic if you run them with a temperature of 0 or a fixed random seed, and your kernel is built to be deterministic, but they're not typically used that way, and will produce different output for identical input.
I never said it is. That's why I qualified my example with the word correct.
> no matter what you do, it gives you the same output
This is not deterministic. This is determined. I think this is the confusion I was pointing out.
>> Deterministic means that if it is accurate/correct once, it will continue to be in future runs (unless the correct answer changes; a stopped clock is deterministic).
The bit in the parenthesis, I am trying to argue, is nonsense. If the correct answer changes, the system is not accurate or correct to begin with so the point is moot. Correcting the system will make it accurate. A stopped clock is not deterministic, it's determined. As a time indicator, a stopped clock is not a correct, accurate or deterministic model at all under any possible interpretation.
You pretty clearly think determinism and correctness are related, otherwise why wouldn't a stopped clock be deterministic?
Determinism is about the behavior of a system. Correctness is also about the purpose of a system. A system can have deterministic behavior while being completely unfit for its purpose. And depending on its purpose, it can be fit for purpose while being nondeterministic.
You still seem to see correctness as a prerequisite for determinstic. I’m open to that idea but I really don’t think it’s the case.
I build a box. It has an LCD display. It has a button labeled “what time is it”. You push the button and it always shows “10:43am”. This is a deterministic system.
That depends. If the problem has been solved before and the answer is known and it is in the corpus, then it can give you the correct answer without actually executing any code.
Is it not generally true? If the information (i.e. problem and its answer) exists in the model's training corpus, then LLMs can provide the correct answer without directly executing anything.
Ask it what the capital of France is, and it will tell you it is Paris. Same with "how do I reverse a string in Python", or whatever problem you have at hand that needs solving (sans searching capability, which makes things more complicated).
So does not the problem need to be unique if you want to be able to claim with certainty it indeed has been executed? I am not sure how you account for the searching capability, and I am not excluding the possibility of having access to execution tools, pretty sure they do.
These two comments stand out to me as inappropriate (directed at OBS).
> keeping up with runtime updates is one of the most basic expectations of a maintainer, and I suspect it's a sign there may be other problems as well.
> I won't mince words: allowing the runtime to go EOL is unacceptable and indicates terrible maintainership.
I don't use Fedora but I do use OBS… on Mac, because OBS is hands-down the most popular application for streaming on any platform. It's crazy that OBS works great on Mac, works great on Windows, works great on Linux if using the OBS Flatpak, and when the Fedora-packaged-flatpak breaks and this Fedora guy starts saying that this is indicative that "there may be other problems".
If OBS isn't good enough for Fedora to ship a working version, then show me the streaming software that is.
"If OBS isn't good enough for Fedora" - fanboyism is never good. If OBS has issues in development then what? What would you do if it stops updating Qt permanently? Think not let emotions act.
"works great" - doesn't mean it is secure.
You can write application that works great and is swiss cheese from security standpoint.
You can write secure application that works like nightmare.
"inappropriate" - why?
If it is statement of fact then it can not be inappropriate.
Also mind you OBS blocked the issue about fact that they use EOL qt on github - this does not look to me as good project.
"the Fedora-packaged-flatpak breaks" - is it broken? Because no one even speaks about real state of package! Or by "broken" you mean - does not have functionality I want! Or it uses Qt version which breaks the application!
Because In first case that not breakage - that's loss of functionality and if motivated by legal reasons - I can understand (not approve since US software patents are from my perspective idiocy), if motivated by security I wholeheartedly approve - because you are shooting messenger(fedora) of bad news(OBS bad practices) here.
In second - Qt is broken so send regards to them and their policy:
Update it so often to make GPL/LGPL version as miserable as possible.
Which they then use to sell companies the LTS versions under proprietary license.
I agree with breaking (it is good feedback about software state) to modernize dependencies - but then again I'm using Arch so…
Not really and I wasted my time browsing a massive thread of non-sense where people were arguing for and against "Fedora Flatpack is better than Flathub" software and vice-versa.
People make a lot as their side projects through the easy parsing, what software is being made with them? (besides the usual hacker news backend response)
> The population, which was around 900 after World War II, dropped to 13 in 2017. At the same time, the number of cats was in the hundreds. Aging residents felt there were too many of them and not enough people to look after them. The Aoshima Cat Protection Society, therefore, recommended that every cat on the island be spayed or neutered in order to gradually reduce the population. The sterilizations took place in 2018.
Is there anybody local that wants the cats to stay? Sounds to me like the program is working as intended. Sure tourists love it, but if there's nobody local to look after the animals welfare, it should end the way it's been planned it seems to me.
>Is there anybody local that wants the cats to stay?
There are five locals. One is the "cat mama" mentioned, who presumably likes the cats but realizes their fate without her is grim. The other four probably don't care that much.
> Sounds to me like the program is working as intended. Sure tourists love it, but if there's nobody local to look around, it should end the way it's been planned it seems to me.
Are we talking about the country of Japan or cat island here? The declining cats of cat island is a wonderful metaphor for the declining population of Japan island.
Makes no sense to me really. Sure the population will decline and maybe in 100 or 200 years Japan's population will recover. Why go through that mess when you can just have 100 million non-Japanese people slide in, just give free citizenship to anyone that has a kid on Japanese soil and don't enforce immigration law, simple as.
I suppose they can Disney-fy it, hire people to live/work there as caretakers, but is that what the area wants/needs?
I remember reading an article about Prague, where the author felt like it's been turned to Disneyland: because of the tourism boom, the city got littered with tourist gift shops and eateries.
we could see cats split into prey and predator, or them develop great fishing abilities or bird lure and climbing capabilities, life finds a way to make many out of few opportunities .
> I'm curious to know if "newly discovered" species existed 20 years ago and were actually just discovered,
That's exactly it, these wasps existed previously and were just discovered to be distinct from other wasps. Speciation tends to take a very long time (on the order of hundreds of thousands to millions of years) or much shorter if there's a strong enough pressure (e.g. something drastically alters an ecosystem and opens up a lot of new niches for a species to radiate into) but still on the order of tens of thousands of years, see [1] for a great example. This of course depends on generation time (evolution only happens to populations, not individuals), so we see quite rapid evolution in things with short generation times like bacteria.
For invetebrates like small wasps like this one, it's typically taking the time to sit down and actually identifying them, some species are quite cryptic and it's only obscure or small morphological features that can be used to separate them by eye, and requires genetic analysis to compare and confirm that it's a new species.
> Examining old hosts that have died and been preserved and seeing if the 'new species' exists there maybe?
I have an entomologist friend and yes, that does happen. There are probably countless new species that have specimens in museums and universities right now that just haven't been properly analysed
Very skeptical. I would also like more detail in the article, what was the DNA collected from? Hair? Saliva? How was it stored, collected and tested? Who sent it in? This website has a bit of detail around what can be tested: https://warwick.ac.uk/fac/sci/lifesci/research/archaeobotany...
The immune system does this all the time. They're called natural killer cells (NK Cells) and they regularly target cells that have become cancerous or infected.
Well yes, NK cells do that when we're infected with a malignant infection.
For the last hundred years, or so, vaccines have been non-malignant, so we've never had to question such things.
However, mRNA vaccines are an entirely novel class of technology, so it's appropriate for us question:
a) how long mRNA vaccines persist in the body?
b) how discriminating are mRNA vaccines in which cells they impact?
c) if any testing was done on the side effects of accidental intravenuous injection of mRNA vaccines? (given that there are plenty of blood vessels in muscle tissue)
d) if any longitudinal studies have been done on mRNA vaccination of humans?
e) if any comprehensive studies have been done on pregnant or nursing women injected with mRNA vaccines, and the impact on their infants?
For those confused by the dangling modifier, he means the combined act of cutting the trees and selling the wood is restrictive. Also, a terrible analogy since trees are an exhaustible resource within a meaningful time range.
