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

BigQuery has that and I've been loving using it since they introduced it


> Limited training data, as compared to kanguages like Python or javascript.

I use my own APL to build neural networks. This is probably the correct answer, and inline with my experience as well.

I changed the semantics and definition of a bunch of functions and none of the coding LLMs out there can even approach writing semidecent APL.


Maybe read the paper first?

> This study asked whether Large Language Models (LLMs) understand sentences in the minimal sense of representing “who did what to whom”. In Experiment 1, we found that the overall geometry of LLM distributed activity patterns failed to capture this information: similaritiesbetween sentences reflected whether they shared syntax more than whether they shared thematic role assignments. Human judgments, in contrast, were strongly driven by this aspect of meaning.

> In Experiment 2, we found limited evidence that thematic role information was available even in a subset of hidden units. Whereas activity patterns in subsets of hidden units often allowed for significant classification of whether sentence pairs had shared vs. opposite thematic role assignments, the effect sizes were small; even the best-performing case appeared to lag behind humans, and its representation of thematic roles did not seem robust across syntactic structures.

> However, thematic role information was reliably available in a large number of attention heads, demonstrating LLMs have the capacity to extract thematic role information. In some cases, information present in attention heads descriptively exceeded human performance.


I wrote a short story ten years ago making fun of this concept: https://blog.chewxy.com/2014/05/20/the-long-term-plan/


Barry Jay's got an upcoming paper at PEPM regarding typed tree calculus. Good read too.


I googled that, and I found that it is a sister conference of POPL: https://popl25.sigplan.org/details/pepm-2025-papers/5/Typed-...

I was inclined to think that Barry Jay is some kind of crackpot


See the Coq proofs in https://github.com/barry-jay-personal/tree-calculus/ for the pre-typed tree content as well.


It's a good thing there are Coq proofs. However, I would still like to see the paper though. A paper is supposed to be more than just a syntactic theory. Typically there is some motivation included as well.


I'm working on my scifi novel. I had started writing it when LLMs started taking off - I had been doing AI for two decades and I was well-placed to be in a good position to profit with the rise of LLMs, but I ended up gaining nothing much and I was depressed about it - so I started writing instead. Been picking at it for about a year before befriending an editor who encouraged me to keep writing. He's helped me developmentally edit it to a point I am now ready to work on my second draft.

It's a hard scifi novel with mild existential horror tones that is borne mostly of maths jokes. At one point the main character tries to escape the matrix (reality). But the matrix is defective, so the best way out was to orthogonalize the subspace and reduce the matrix to its eigenbasis instead. Most of the scenes are based on similar maths jokes.

Tentative name is Diagonalization of the Meta (I had previously called it The Metaverse).


What is your GTM strategy?


At this point I'm writing mostly for myself. GTM strategies for novels... that's an interesting way to think about things. I've not thought about it just yet. Happy to hear if you have any ideas tho.


good title!


Thanks :)



Is there a no commercial use license?


Not really, this is actually pretty readable


I told a variant of the original Little Mermaid story as part of a school outreach program. The kids came to the conclusion that God wasn't a fair being because he didn't give mermaids souls. I walked away satisfied that my little counterprogramming against catholic school indoctrination might have worked. I wasn't invited back (at least for school year 2024).


In some novel, the author discussed Abraham's sacrifice of Isaac [0] as not a test of Abraham by God, but a test of God by Abraham.

As in, 'I am about to murder my only son on your orders. If you are indeed the kind of god who would order me to do such a thing, then we'll see where that leaves us...'

That interpretation always struck me as truer to Old Testament tone.

[0] https://biblehub.com/kjv/genesis/22.htm


At the time, child sacrifice was apparently common, enough that if a country was in trouble, the populace would demand the king sacrifice his kid to save the country (even shown in scripture … see 2 kings 3:27 though later in time). This was a very _public_ display that this God does not want that.

In short, it wasn’t really a test of either one, it was a public declaration that child sacrifice is bad.


Didn't work out very well for Stanis Baratheon either.


Sounds like Dan Simmons' Hyperion and Fall of Hyperion. I think that particular bit was in the second book, but Sol spent a lot of time grappling with Abraham in both.


That sounds correct and in keeping with the themes!


I don't know. If memory serves life was pretty cheap in the old testament with millions being murdered and everyone(?) killed if you count the flood.


Within the context of the narrative, Isaac's importance to Abraham was practically infinite.


Weren't those non-believers, though?

Old Testament God is pretty firm on that line. :D


I'm not surprised the school didn't invite you back. Was the school outreach programme organised by your employer?


Leaving the classroom, I tip my fedora and chuckle to myself. As I smile at my own cleverness I wonder how much karma this story is going to get when I post it on the atheism subreddit later.


You made the common mistake of assuming God was/is a being: https://nwcatholic.org/voices/bishop-robert-barron/who-god-i...

Setting up a strawman for the kids would be par for the course though.


I wouldn't blame anyone for assuming God is a being. It's hard to reconcile the idea that God is both an abstract entity, like a force in the universe, but it also can become fully human as Jesus Christ.


That framing is a bit of a stretch given the widespread tendency of the religious to anthropomorphize God in terms like having human-grokkable preferences and communicating them to us.

I'd say that argument has itself preemptively "retreat[ed] onto ever-shrinking intellectual turf. Defining God as something akin to the entire existence of the universe is something that essentially cannot be proved or disproved. Stick to that definition strictly, and yes there is nothing that an atheist can take logical issue with. But that strict definition also yields no conclusions/advice/insight either, so it's not very interesting. Hence seemingly no one ever being able to adopt such a definition and actually stick to it.


Any metaphysical framing of ‘why existence’ is a bit of a stretch and can never be proved or disproved. I’ve also wondered whether a logical atheist would care if they were logical considering time is zero sum :) Also, these ideas are harder to grok in the modern mindset of reductionism (also unprovable), but this conception of God and being is millenia old.


I mean the framing of an abstract non-entity God is a stretch from how basically everyone actually invokes God. Sure, that conception of God has been around a long time - however I've yet to come across any religion that sticks to that conception. Instead it's generally used as part of a Motte and Bailey setup - such an abstract conception of God cannot be disproved, and so one has to agree that such a God may exist. But then having established that, the general feeling that there is some kind of higher power is used to give weight to a whole bunch of assertions of what a completely different conception of a higher power supposedly wants us to do.


Maybe in some circles, but the abstract idea of God being love (assuming one has faith that consciousness and free will exists in reality) go back millennia. The abstractions predate the Simpson’s ‘beard in the sky’ by a few years ;)


there's also a good paper - Can a neuroscientist understand a microprocessor - by Kording's lab which is also an excellent read.


...but there's nothing more obnoxious than a physicist first encountering a new subject.

https://xkcd.com/793/


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

Search: