wrsh07

Joined 12 years ago

Karma

2,466

Submissions

1,037

Member since

Jan 15

View on Hacker News

Recent submissions

Comment · 16 days ago

If you can increase your budget to 30k, you could consider a Slate EV.

Electric cars are still an underdeveloped form factor, so it makes sense that you can't get a car big enough for many passengers that is cheap enough for your budget (the battery needs to scale for passengers/weight, so the market so far is tiny cars or big expensive cars. The middle comes as the tech develops)

Comment · 23 days ago

It's not just that people have worked on it for a long time, it's that there are dozens of problems that are known to be 3sum hard and people have tried to make those as efficient as possible too!

Eg here's a paper with a bunch https://people.csail.mit.edu/virgi/6.1420/papers/gajovermars...

So every grad student who has thought about geometry problems will have spent more time than they're proud to admit thinking about better algorithms only to realize it's 3sum hard.

It's not just "people have thought about one problem a bunch" - the whole class has been wailed on to no avail

Comment · 26 days ago

Unfortunately, the guidelines generally prohibit it:

> Don't be snarky

https://news.ycombinator.com/newsguidelines.html

Comment · 26 days ago

Nobody thought this was possible.

3sum hard was colloquially considered to be >= n^2

It's an absolutely unbelievable result! (Personally, this is more meaningful to me than Navier Stokes and feels more surprising - not that an agent did it but the result itself is extremely surprising!)

Comment · 26 days ago

Evidence for vs compatible with are different things, and agreed!

I don't think it's particularly controversial to say "written language is a meaningful technology advancement for a civilization", though!

Comment · 26 days ago

It seems pretty clear that this is a serious scientific endeavor. I think it's fine for you to be skeptical, but it doesn't feel particularly "in the hacker news spirit" to sarcastically write it off like this

Comment · 26 days ago

Language and written language are themselves technologies that change what individuals can do (and thus impact their societies)

There is a well-known story about this discussed in Radiolab [2010] https://radiolab.org/podcast/91725-words/transcript

Here's ChatGPT summarizing one section:

> Younger signers had developed a richer vocabulary for mental states—concepts like believe, remember, forget, understand. They were much better than the original older signers at tasks requiring them to reason about what another person thinks or knows. Later, as the older signers picked up those mental-state terms from younger people, some of them improved on the same tests. Senghas describes it as language giving them new insight into other people’s thinking and motivations.

I see several likely stories:

* Much animal communication is practical, descriptive of the present (nothing abstract, or very few abstract concepts)

* If a species could expand its communication, it would give it more powerful tools and allow for more meaningful conversation with humans

* If such a thing is possible, it will be extremely disruptive to that species

I think if you could communicate about abstract ideas with your dog, you could exchange bedtime stories. However, I don't think such a thing is possible today, and it's not clear what species (if any, besides humans) will, in fact, be able to learn such things

Comment · 1 month ago

I'm confused why there's no attribution or reference to Tobi's (which in turn referenced the cursor git post)

Attribution is not just a problem for LLMs in mathematics!

Tobi is where I first heard the concept: https://github.com/tobi/walgit

And he credits this Cursor post for giving him the ideas and says it's worth reading: https://cursor.com/blog/git-at-any-scale

Note: in related work, the East River Source Control team has posted about their vision for git as well: https://ersc.io/blog/what-comes-after-git

Comment · 1 month ago

You know, the first time you navigate somewhere (if you don't already have perfect directions) will probably be the longest route you'll ever take to get there

For Earth, the proof presented for NS is just our first attempt navigating from our previously known facts to the proof.

I expect we will be able to shorten it dramatically (most likely with human and AI insights), but I don't think we should read too much into the length. If you want a similar point of comparison, see the original proof (by humans) of Fermat's last theorem. It has been shortened significantly. This is normal.

Comment · 1 month ago

The rule of thumb is simple: if most of a user or author's writing is consistently human-generated, then I think we are very happy giving them the benefit of the doubt if one article or snippet flags the actually good ai detector

Unfortunately, many of the largest voices against pangram simply don't like it because it gives their lies less credibility.

I don't mind reading llm writing, indeed, I read more llm writing per day than most. But if you're using LLMs to increase the level of slop (blog posts, comments, tweets, etc) then we should call people out.

It's a colossal waste of everyone's time and attention, and we should be mad about it.

Late edit: if people are actually writing and it's consistently flagged by pangram (this is a statement I have not yet seen validated), the pangram folks are extremely proactive and excited to understand what's going on. This notion that somehow the authors of slop are victims is complete nonsense.