sva_

model card https://news.ycombinator.com/item?id=49537354 (doesn't 404)

Yes something like that is what I did, so I had my 5 hour reset window to be at 2 hours so I could work. But anthropic reset it so it went back to 5 hours.

So I need to center a div

Great, my usage reset is in 10 hours ...

And my 5 hour window was due to be reset in 2 hours (barely used), now its in 5 hours - so this reset effectively gives me 1 less 5 hour reset for this weekly cycle.

I am not sure how someone would run an LLM on my computer to look for exploits as the source code is only available on my machine?

Even if you introduce bugs in your UI (which I think is not very likely if you have a basic understanding of your system), the chance that someone would exploit software that literally only runs on your own machine seems extremely unlikely to me.

I've been using arch for over 10 years btw.

I think this is more about the UI, rather than the install. I haven't tried it myself though.

I think nowadays using quickshell anyone who is so inclined can vibecode their own UI though. I recently made the switch to Wayland/hyprland and rebuilt my polybar on quickshell, even adding widgets that allow getting system info/fine grained system control (interactive Bluetooth, WiFi, Volume, Brightness etc).

Glad to see you two read only the title, not the article; that fulfills the cliche.

And in your list you forgot taking a lecture in category theory.

You can add several remotes to your git, and I'd recommend you do so.

Apple forcing users to use their own browser/browsing engine doesn't disprove my argument IMO, virtually nobody outside the apple ecosystem uses Safari, and outside the apple ecosystem is something between 80-90% of internet users.

Hot take: Google keeps Mozilla/Firefox alive through the default search engine placement (which makes Mozilla millions each year) so they don't get designated as a monopoly with their browser.

I think they also have the problem of having given away their pro subscription to 10s or 100s of millions of students worldwide. They're tightening down on that now, and I have a feeling that this goes into them not releasing a larger model.

In 95% of cases it is because you had a large context for which the cache expired.

I'd say the EU is the region with the highest acceleration of climate scientists, hence they find "the highest acceleration of most climate parameters".

The title is editorialized. It is "Keep AI Out of Your (Obsidian) Vault".

For the Original title, I wish they added the qualifier "generative" for AI. You may well use e.g. embeddings to semantically search through your notes, and there are only gains in that.

Having an AI write notes for you is however of course completely pointless. The act of writing notes is 80% of the reward.

The EU is not "most at risk" compared to eg south Asia, the middle east, parts of Africa, etc; and this idea that AMOC collapse will turn Europe into a glacier is a blatant exaggeration that does not root itself in reality.

Even if true, I wonder if it really helps if I basically need to hire a lawyer ($$$) to find out which regulations may or may not apply, and if I'll get myself into trouble.

It is not a coincidence that here in Germany there are whole law firms specialized in suing small business owners (Abmahnkanzleien.)

I think for a lot of people the tradeoff might not be worth it.

> Content that is entirely generated [...]

Take note of the qualifier entirely. If you're working with an agent steering it to produce the results you want, it would be an entirely different story.

I did not have that experience at all, ordered there dozens of times.

This "low power mode" is a myth, the brain pulls a pretty constant amount of power (about 20% of metabolic energy), and most power isn't consumed by the action potential/neurons firing, but rather in keeping the neuron ready to fire.

And the regions involved in the default mode network are actually some of the metabolically most active regions in the brain, involved in memory consolidation, future simulation, theory of mind, and self-narrative; among other things.

It is a proprietary eponym.

> The submission process is thorough, but achievable: as a test, I successfully managed to submit my own recent formalization [...]

I find this kind of endearing, how he almost makes it sounds like "If even I can do it, you can too!", disregarding he's probably the most prolific mathematician currently alive.

Some years ago I suggested it might be interesting to have a kind of blockchain of formally verifiable mathematical proofs, but was quickly shutdown by people telling me it was a useless endeavor due to Goedels incompleteness theorems.

I think the larger issue of computationally creating a library of math proofs is still that one might come up with an infinite amount of useless theorems that are trivial to prove, but I suppose this registry is manually vetted. Theres a strong inductive bias in maths in that humans still decide what axiomatic systems, theorems, definitions etc are interesting to us.

But I think the method of proving things will soon go out style as an endeavor to devote perhaps years of your life to.

> enabling massive clusters and models with more than 50 trillion parameters

Yeah I also had an EVGA 1080 Ti die on me, and got a better GPU as upgrade. Their warranty was worth the premium price for sure.

Overall I have made very good experiences with long hardware warranties. Soldered RAM on my ThinkPad died once and I got an entire new mainboard after almost 3 years (it died 2 months before warranty expired.) New one still works today but the hardware is just too old.

Had the EVGA 1000w fail on me. It did not fail dramatically though, rather it refused to output high power. Got a replacement under warranty

Sounds like you're burnt out. Maybe its time to touch some grass and think about what is really important and meaningful to you.

About 150 years ago, Nietzsche wrote

> God is dead. God remains dead. And we have killed him. How shall we comfort ourselves, the murderers of all murderers?

about the secularization of society. Perhaps something similar is happening to knowledge work now, but I think it was never a great idea to derive all your life's meaning from it anyway.

> The very fact that there is generally no "best next token" with 100% certainty

This is not entirely accurate. Sure, there's never a token with 100% certainty, but there are often tokens with 99.9% probability, but this technique of course does not change how such a token is sampled.

Claude Code works for me (Opus 5).

The Web might be better off if those offenders aren't part of the web anymore.

It's turtles all the way down