Hacker Newsnew | past | comments | ask | show | jobs | submit | progval's commentslogin

bluebarbet meant it should be "Gen Z and its parents".

Both "Gen Z and their…" and "Gen Z and its…" are fine. I personally prefer to treat a group as a plural.

"The class tossed their graduation caps in the air" strikes me as less clumsy than its alternatives.


Well, considering Gen Z is a group of people who has parents, their works just fine. "my Mom" "my Dad" ... parents are usually referred to as entities the child has. And Gen Z is a plurality of people.

No, it shouldn't. It's perfectly valid to refer to the members of a group rather than the group itself.

Do they count as export if the owner doesn't change?

If central banks sell their gold in the US and buy new gold abroad (to save themselves physically transporting the gold across the ocean[0]), I suspect that the demand this creates for gold abroad will ultimately cause US gold exports to rise, too.

[0]: See the link in https://news.ycombinator.com/item?id=50134172


Selling it in place neatly sidesteps suspicion about how secure the NY gold really is and if it's all really still there, too.

If this suspicion existed then you'd see a spread between (gold price in the US) and (gold price outside the US)

When Trump screwed with the tariffs we saw exactly that. Arbitrarily messing with tariffs is one of the risks that causes people to think keeping gold in the US is not entirely safe.

Not necessarily. If relations are good, people are willing to look the other way a lot.

Either way, you aren't left holding the bag, you're giving the bag to speculators.

Do they count as export if the owner doesn't change?

I don't know. Does it count as export if it's not moved an inch?


For the inference APIs, which need to bill users. It's unclear why OpenCode's servers (presumably) had access to the API credentials.

They function similar to openrouter BYOK.

Because people cut and pasted them in.

Maybe they support cloud agents.


Weechat does it: https://weechat.org/files/doc/stable/weechat_quickstart.en.h... I don't use this feature, though.

However I like that I can see all the channel with unread messages at once, and can switch between them quickly. On Discord if I'm chatting in three channels spread across two servers then switching takes two clicks.


Emojis and reacts are the only extensions that can cause people to have a meaningfully different view. It's part of why it took so long to ratify replies, and reacts still are not ratified.

We're also discussing ways for servers to provide a fallback to clients which don't support them.


> We're also discussing ways for servers to provide a fallback to clients which don't support them.

Uh. I feel like this should be something that's done before it's ratified. What if you find out that there is no satisfactory way?


It is the protocol's fault, and there is a solution at the protocol level: https://ircv3.net/specs/extensions/chathistory . So now, you can get messages while offline when using a modern server (eg. https://ergo.chat/) with a modern client (eg. https://codeberg.org/emersion/goguma)

Ergo looks really cool, will have to take a look at that. Thanks for sharing.

Oh thanks for the note about ergo.chat, need to have a play with this!

Not every optimisation is the same. LLMs can optimize code that is behind known practices, but won't usually invent new data structures or algorithms to solve a unique problem.

While it's unlikely for LLMs to implement patterns not present in their training, current training techniques absolutely can produce practices and data structures and algorithms that humans have never done before. Modern LLMs are not just trained on text. They are trained on computer use, including - at a high level - solving programming problems with code. They are not just rewarded for producing the most likely next token. They are rewarded for producing tokens that solve problems. Patterns and techniques can emerge from this that are not present in any human behavior.

Yes, but in some fashion, all the outputs are mathematical convolutions of the training, plus whatever is packed into context.

They are rewarded for code compiling, thats about it. Thats why the code is terrible. Ive yet to look at a project made by an llm whos user swore this "frontier model is the one", where the code wasnt utter garbage. The heavy llm users dont care as long as it compiles.

I'd be interested at which problem you passed to an (which?) LLM that they weren't able to solve?

It's meant to protect apps against their own bugs. It's an improvement to be enabled by default and disabled for the few apps where it's an issue.


It's a major plot point in one of the books in the Bobiverse series, though it's a civilisation humanity meets instead of humanity itself (not saying anything more to avoid spoilers).


Consider applying for YC's Winter 2027 batch! Applications are open till November 2.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: