Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
geremiiah
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
91.
▲
by
geremiiah
9mo ago
There's the Citroen Nemo. It's a more compact version and I believe closer in size to the older panel vans. The new Berlingos are wider and taller than they used to be.
92.
▲
by
geremiiah
9mo ago
> We need our economic incentives to be aligned with long term growth Isn't that asking for more TSLAs? I don't mean Tesla the company. I mean TSLA the stock. People hold it because they want to bet long on the future.
93.
▲
by
geremiiah
9mo ago
https://www.euronews.com/my-europe/2026/01/09/eu-rejects-vio... > The crackdown on protesters in Iran has prompted an outcry among European leaders, with High Representative Kaja Kallas denouncing Teh
94.
▲
by
geremiiah
9mo ago
The article says basically nothing besides, "people are protesting in Iran". I found this survey, no endorsement because I have no idea about the source, but it gives a better picture of where the Iranian people stand: https:&#x
95.
▲
by
geremiiah
10mo ago
Capitalism and democratic systems don't work without growth. Just look at the consequences of the economy stagnating in western Europe and what that is causing on a societal and political level. No growth means no jobs for new grads, n
96.
▲
by
geremiiah
10mo ago
So you're basically talking about HDI. That's one metric. But I think equally important is also to consider the rote tech capabilities. That India still has a lot of people practicing susbistence farming and that China still has p
97.
▲
by
geremiiah
10mo ago
Ironically the reason Karpathy's is better is because he livecoded it and I can be sure it's not some LLM vomit. Unfortunately, we are now indundated with newbies posting their projects/tutorials/guides in the hopes that
98.
▲
by
geremiiah
10mo ago
The risk for Tinygrad is that PyTorch will create a new backend for Inductor, plug in their AMD codegen stuff and walala, PyTorch still king. I mean, they could have easily just taken that route themselves instead of bothering with a new ML
99.
▲
by
geremiiah
10mo ago
In a parallel universe RMS read Ayn Rand, joined her cult of selfish greed and the world is running a closed source UNIX clone developed by JLU (Just Like Unix), the $5T corporation funded by him. All of this to say, it's amazing how t
100.
▲
by
geremiiah
10mo ago
OK, suppose the whole world used a trustless digital cash. Now you cannot do monetary policy. Now booms are booms and busts are busts and we are back to the early 1900s. Great. Actually what would likely happen is that people would be incen
101.
▲
by
geremiiah
1y ago
Check her Twitter. She supports Trump.
102.
▲
by
geremiiah
1y ago
100%. The biggest economic problem plaguing Germany and most of the EU is that all the capital is in the hands of incompetent old farts, so called, "business leaders", and the economic class system is way more hierarchical and old
103.
▲
by
geremiiah
1y ago
Sorry, but lol. Germany even makes you pay to be able to piss in a toilet. Even public libraries aren't free! And let's not mention the health care system, which literally has two tiers, one for the poor and one for the rich, with
104.
▲
by
geremiiah
1y ago
The communist parties have very low ratings and never make it into coalition governments so they are not a threat. The Greens also have fairly low ratings and despite the image, they tend to be elitist neoliberal types. None of these peopl
105.
▲
by
geremiiah
1y ago
Comments like these make me realize most people have no clue what science is really like.
106.
▲
by
geremiiah
1y ago
Thanks. Certainly lots of work seems to have gone into this, but if they had any good results they would be front and centre. Skimming the paper, I can only see model architectural details and some links they try to make to biology. That ma
107.
▲
by
geremiiah
1y ago
> industry donors Which ones? The meat industry or the soy farmers?
108.
▲
by
geremiiah
1y ago
But generative models are always going to seem like they are doing ok. That's how they work. They are good at imitating and producing misleading demos.
109.
▲
by
geremiiah
1y ago
Their representation is simpler, just a transformer. That means you can just plug in all the theory and tools that have been developed specifically for transformers, most importantly you can scale the model easier. But more than that, I thi
110.
▲
by
geremiiah
1y ago
They just want to feel like they are perpetual victims. It appeals to them and it justifies their cruelty towards others. I guess this is a major difference between the far-left and the far-right. The far-left go out of their way to join th
111.
▲
by
geremiiah
1y ago
In British English, "middle class" refers to the well off professional classes or merchant traders. In American English, if I understand correctly, everyone who works is considered middle class.
112.
▲
by
geremiiah
1y ago
Kids who used to study electrical engineering, or mechanical engineering or go into finance and accounting, decided all to major in computer science instead, because if you are going to be an office drone, why not be an office drone in the
113.
▲
by
geremiiah
1y ago
It's more akin to KYC and AML laws. I've heard many horror stories of bank/brokerage accounts being frozen despite all deposits being legitimate, not to mention not being able to transfer large amounts or not even being allow
114.
▲
by
geremiiah
1y ago
4chan isn't rightwing and it never was, but since 2016 and especially since 2020 the userbase has been moving to the left, because at its core 4chan is always counter culture.
115.
▲
by
geremiiah
1y ago
Simple models shouldn't be applied without nuance (and that's why I get really annoyed at "engineers" on HN who think they know jack shit about economics because they took econ 101). The idea is sound, but the real world
116.
▲
by
geremiiah
1y ago
It's kinda funny that Americans look at the German economy and think they want to be in that situation. You would rather have an economy based on manufacturing cars, rather than one based on leading in tech? You sure?
117.
▲
by
geremiiah
1y ago
AI helps you cook code faster, but you still need to have a good understanding of the code. Just because the writing part is done quicker doesn't mean a developer can now shoulder more responsibility. This will only lead to burn out, b
118.
▲
by
geremiiah
1y ago
It's more widespread than PINNs. PINNs have been widely known to be rubbish a long time ago. But the general failure of using ML for physics problems is much more widespread. Where ML generally shines is either when you have relatively
119.
▲
by
geremiiah
1y ago
LLMs are not involved anywhere. You start with some data. Either simulation data or experimental data. Then you train a model to either learn a time evolution operator or a force field. Then you apply it to more input data, and you visualiz
120.
▲
by
geremiiah
1y ago
I don't want to generalize because I do not know how widespread this pattern is, but my job has me hopping between a few HPC centers around Germany, and a pattern I notice is that, a lot of these places are chuck full of reject physici
More ›