For the best experience on desktop, install the Chrome extension to track your reading on news.ycombinator.com
Hacker Newsnew | past | comments | ask | show | jobs | submit | history | fcpk's commentsregister

it's high time actual price starts to hit subscriptions rather than VC subsidized first hit is free stuff

quite brilliant though don't expect any easy therapies out of it yet. there is no known small fragment(ie manufacturable peptides) that bind to PLXNA1 , and the terminal fragment is about 400 amino acids, way more than what can be produced with solid peptide chemistry (100-150 amino).


something I have been wondering about is doing regressive layer specific quantization based on large test sets. ie reduce very specifically layers that don't improve general quality.


This is a thing! For example, https://arxiv.org/abs/2511.06516


that's brilliant, I wonder why we haven't seen much use of it to do very heavy quantization


This is a very well established idea. It's called dynamic quantization. Vary the quantization bit-width (or skip quantization altogether) on a layer by layer basis, using a calibration dataset.

EvoPress is the first time that comes to my mind, when I think of dynamic quantization.

https://arxiv.org/abs/2410.14649


I've experimented with this with diffusion models with a safetensors - gguf tool I wrote. even with relatively few sample images (~10k, still enough to keep my 3090 spinning for days straight) the benefits are quite noticeable - a smaller file with overall better results.


I mean it works... but it's really ghetto. You have to handle username collisions(or enforce unique usernames). IPv4 should be non free, and that'd cover the costs...


Iran International (Persian: ایران اینترنشنال, romanized: Irān Internašnāl) is a Persian-language satellite television channel and multilingual digital news operation based in London, United Kingdom. Established in May 2017 and reportedly linked to Saudi Arabia.

from wiki

this is just a mouthpiece of the Saudis probably condoned by MI6. take it with a grain of salt.


and where's the result? LOC as a side a measure of success is typical for the "omg LLM are amazing and can do it all phase" but once you enter the "actually shipping products people want with human complexity and experience meltdowns" it's usually different....


Well at the current trajectory I'd expect him to release his own OS or something by end of July, his own AWS competitor by October and to close YC applications indefinitely at the end of the year.

But for now I'd be fine with him making his repos public.


this begs the question... how are we seeing life patterns appear from that?


Raises.


The fact there is no mention of what were the bugs is a little odd. It'd really be nice to see if this is a "weird never happening edge case" or actual issues. LLMs have uncanny abilities to identify failure patterns that it has seen before, but they are not necessarily meaningful.



The fact that some of the Claude-discovered bugs were quite severe is also a little more than something to brush off as "yeah, LLM, whatever". The lists reads quite meaningful to me, but I'm not a security expert anyways.


Here's a write-up for one of the bugs they found: https://red.anthropic.com/2026/exploit/


I’m guessing it might be some of these: https://www.mozilla.org/en-US/security/advisories/mfsa2026-1...


Yeah, the ones reported by Evyatar Ben Asher et al.


I correctly misread that as “et AI”.


we can put that one next to the Weird AI Yankovic music generator.


“et AI, Brutus!"


Yon Claude has a lean and hungry look.


An LLM by any other name would hallucinate the same


Anyone still reading down here will appreciate this https://bsky.app/profile/simeonthefool.bsky.social/post/3kbk...


Hang on, someone downvoted me for a horrific pun? GOOD.


I upvoted, so maybe that restored the balance.


Out, out, vile upvote.


He computes too much.


Indeed, without it looks like a fluffy marketing piece.


And now that you know that it isn't, do you feel differently about the logic you used to write this comment?


i am curious, what are you hoping to get out of this comment? will you feel better if they say yes? what is your plan if they say no?


I genuinely want to understand how they arrived at the claim that this was a fluffy marketing piece. Like, if you said on a different thread, "the Linux kernel is probably mostly written in Pascal", I would really want to understand how it was you got to that idea.


> what are you hoping to get out of this comment?

Rando here. It gives a signal on the account’s other comments, as well as the value of the original comment (as a hypothesis, albeit a wrong one, versus blind raging).


>"It gives a signal on the account's other comments,"

fair enough. i typically use karma as a rough proxy for that, especially when the user has a lot of it (like, in this case, where the poster is #17 on the leaderboard with 100,000+ karma). you dont get that much karma if you are consistently posting bad takes.

>as well as the value of the original comment (as a hypothesis, albeit a wrong one, versus blind raging).

i dont see, in this case anyways, how or why that distinction would matter or change anything (in this case specifically, what would you change or do differently if it was a hypothesis or simple "raging"?), but im probably just thinking about it incorrectly.


I think a lot of people are overreading this and really all that's happened here is that I was out at a show last night and was really foggy when I woke up and asked a question clumsily. It happens!


yeah, absolutely, i was not intending to start some big inquisition against you or anything.

just like you were genuinely trying to understand where pjmlp was coming from, i was genuinely trying to understand what you would get out of an answer to your question (or, like, what the next reply could even be other than "ok, cool").


Oh, yeah, no, you're fine, this is on me.


> you dont get that much karma if you are consistently posting bad takes.

I wonder how true that is. While this site doesn't have incentivize engagement-maximizing behaviour (posting ragebait) like some other sites do, I would imagine that simply posting more is the best way to accrue karma long-term.


>I would imagine that simply posting more is the best way to accrue karma long-term.

i definitely agree, which is why i use it as a rough proxy rather than ground truth, but i have my doubts that you can casually "post more" your way into the top 20 karma users of all time.


Do I?


I don't know. I'm really asking. I have you bucketed in my head in the cohort of "HN commenters who write lots of assembly", so the mismatch between your prediction and the outcome is just really interesting to me.


this is true for weight loss while in a mostly health range. it isn't true for people with high BMIs. for them the scale is a mostly direct match to health and how they look


I think they are also at risk from the BMI simplification and the goal as presented in the article. The wrong crash diet approach can be raising their % body fat while lowering their BMI. If they weren't weight lifters that's probably putting them in a high risk group that isn't usually identified in studies.


human bodies are incredibly optimized to store energy for bad times and maintain a good balance at the ideal level. this is why the whole "calories in calories out" is technically true but mostly useless for a lot of people as the calories out is a very hard to compute variable. it also ignores entirely the negative effects of muscle loss.

there is a lot of complexity and difficulty in long term weight loss, we are fighting biology hard there. that's why glp1 agonists are having such a success, they allow to fight hormonal homeostasis with proper weapons.

I have yet to see any research showing a way to durably affect the body's set balance, which would be the revolution.

but this research all in all confirms what we know: * high intensity muscle fiber tearing exercise is much better at not affecting metabolic compensating mechanisms * cardio might be good for health and other things, it's very much neutral or possibly negative for weight loss, as cardio does not build muscle mass * pure diet changes are difficult to make sustainable for many. I have seen it first hand where a constant 300 calories deficit a day resulted in weight gain and muscle mass loss despite cardio.


Humans are no different than other mammals.

Look at overweight house cats and dogs. How did they get overweight? How do they lose weight?

Calories in / Calories out is the definition of survival or not for wild mammals.

Calories out is also exceedingly easy to calculate- on any given week if you gain weight the out is less than the in. If you maintain its ahout equal. If you lose weight out is higher than in.

Nothing else matters.


calories in calories out isn’t true at all

as a simple example consider eating 100 calories of vegetables vs 100 calories of pure sugar. your body has to work a lot harder to digest one of those vs the other


A more importantly you’ll still lose weight if you eat 1200 calories of sugar daily if you’re burning 1500 daily.


and you’ll lose even more if those calories are not ultra refined carbs


Sure, but at this point we’re interested in helping people lose weight who just “can’t”.

Fix the major health problem, then refine the details.


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

Search:

HN For You