HN user

finebalance

139 karma
Posts0
Comments44
View on HN
No posts found.

This war was unavoidable btw, it was going to happen sometime this year or next.

Iran was, as per the latest reports I've read, complying with terms and not enriching uranium to weapons-grade or close to weapons-grade. Are there credible reports suggesting otherwise?

I do it more or less this way - except I keep the root intact until the end. It keeps the onion structurally intact until I'm done with the dicing. At which point, the root takes a single chop to lop off, and then the whole thing scatters into tiny, mostly uniform dices. It's quite satisfying.

not that a single user uses a lot of energy for a simple chat.

This seems like a classic tragedy of the commons, no? An individual has a minor impact, but the rationale switching to LLM tools by the collective will likely have a massive impact.

Gemini 2.5 1 year ago

That's ok. AI will kill those off soon enough, and like all winners, rewrite history enough so that that inconvenient theft never happened anyway. It's manifest destiny, or something.

Not a clue.

I'm a decent engineer working as a DS in a consulting firm. In my last two projects, I checked in (or corrected) so much more code than the other two junior DS's in my team, that at the end some 80%-90% of the ML-related stuff had been directly built, corrected or optimized by me. And most of the rest that wasn't, was mostly because it was boilerplate. LLMs were pivotal in this.

And I am only a moderately skilled engineer. I can easily see somebody with more experience and skills doing this to me, and making me nearly redundant.

And in all of this no-one gives a crap about the children, since they rarely spend even a word talking about tracking down the uploaders or creators of said porn (i.e. they say a site "has child porn", but not how fast it is removed, or if the site gives the IP of child-porn uploaders to the police, or anything beyond trying to establish in the viewer a bad site <-> child porn association).

Aren't there legal requirements in place for this stuff, at least in the US?

Collective fucking madness.

It's tiring to be subject to the constant narcissism. The human toll of the incompetence of this government is incalculable. They've set us back decades, if not forever, because I have no clue how our dumb bureaucracy and spineless civil society are going to deal with destruction that climate change is likely to visit on the subcontinent without completely fragmenting.

I swear. But it's not just an issue with the cheap, knockoff stuff: my Bose Soundsport Earbuds often fail, too, with the left one connecting maybe two thirds of the time. It's so infuriating. The only upside is that they were a gift rather than a purchase, so at least I don't have to suffer through the fact that I paid for this crap.

In so many of these papers it seems to be almost a given that the global south is fucked so, holding that constant, they do a cost-benefit analysis on various levels of warming for already (mostly) rich countries.

It's kind of astounding how rapidly the series has disappeared from the mainstream. Couldn't have bungled it up worse had they tried.

I'm Indian, and I think there's a strong case to be made that the Indian avg IQ is likely to be lower than 100. This is nurture though, not nature: a large % of Indians do not get sufficient micronutriants during childhood, and are anemic (40%~) [1], stunted (40%) [2], and much else besides.

[1] https://www.hospitalmanagement.net/comment/anemia-prevalence...

[2] https://www.indiaspend.com/38-of-indian-children-under-4-poo...

It matters, in certain contexts.

For example: a large number of clustering methods boil down to matrix factorization, with variations in the constraints. If you have both domain understanding and a general understanding of what kind of output these variations are likely to result in, you can often narrow down the list of methods you need to try.

Also, interviews.

The point the poster above is trying to make is that given the same amount of data, improvements in technique is leading to significant improvements in accuracy.

An illustrative example comes from the first lesson in fastai's deep learning course: an image classifier that would have been SOTA as late as 2012/13, can be built by the hobbyist in like 30 seconds.

That said, I don't disagree that this is all narrow AI, at best.

I think you've hit the nail on the head here. I live in a country where people don't get enough protein due to religious and cultural dietary restrictions. Once I got my own apartment I upped my protein intake and significantly limited my carbs, and found that the constant cravings I'd had pretty much my entire life decreased substantially, after a short but painful transition period.

The people who are good at this are just so much better at it than the average person. A fast woman, even if she isn't faster than the fastest men, will still be faster than 99.9+% of them.

That's not in doubt. An extremely skilled person will outcompete the field, all other things being equal. Unfortunately, with reference to most sports, all other things aren't equal, which is why good high-school male teams in sports such as soccer can credibly compete with national-level adult female teams.