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

There isn't even deepseek V4.

I'd rather trust LLM arena leaderboard, which puts it on par with sonnet.


LM Arena uses human side by side voting, which limits its applicability to complex tasks.

The ARCPrize leaderboard does have Deepseek V3.2, which only scored 4% on ARC-AGI 2 (while the top models score over 80%). It also Kimi and Qwen, but they also didn't perform well.


Deepseek is a fraction of the cost of western LLM and still just as good. I say it's also related.

I think that plugins are an inherent risk, there is a pop up in obsidian warning the user before enabling them, and it's up to the user to agree or not.

In my opinion, what could have been done is kind of like what mozilla does where it will vet some of the most popular extensions, so that you know there is at least some kind of verification on these extension, and let everything else be wild.

I'm not sure that you can use a.i. to defeat a.i., if an ai is able to spot malware in a code, it can just as well hide it (from itself).


The blog post describes this but there are still manual reviews, similar to what you are asking for. We just need to expose that in the UI.

AI is not used in the review process. The system is primarily based on our open source eslint plugin, with additional dependency and malware scanning

https://github.com/obsidianmd/eslint-plugin


At this point, if you want a laptop, get a mac and be done with it.

Until other manufacturer step up their game, there will be years and years.


Apple was given a free run by Intel's fab issues. I'm hoping Panther Lake laptops together with Dell's CAMM2 will make Linux on amd64 highly competitive with Linux on the M-series, so maybe months and months - not years and years.

Omg. All the worst of Google embedded in a single device.

I swear if I see it in real life I'm going to spray holy water on it.

A.I., data collecting at every level, horrible incoherent ui.

Hit me daddy !


Everyone gets unhealthy CO2 levels if the bedroom doesn't have air extraction inside or you don't leave the door opened.

Does it really matter in the grand scheme of things tho ? I have a captor at home, even when I leave the door opened and CO2 remains low, I don't notice anything different at wakeup.


It can be the difference between between feeling like you're suffocating, not getting enough oxygen to rest enough/sleep well.

I notice a difference if I move between a ventilated room vs congested one. I suppose it depends on what's causing the concentration. If it's human breath, I'll smell freshness. If it's e.g. burning a portable gas heater (common in my part of the world), I'll feel like I'm not inhaling smoke (probably small amounts of CO).

A few years ago, I would sometimes wake up at night and open a window wide, or go open the outside door and stand there for 5-10 minutes.


Sidenote but still, I put ear plugs for 5 years already. It doesn't hurt, you don't feel anything, but it helps tremendously to stay asleep. Better than trying to fix your environment.

I even heard of people going to sleep with airpods pro in their ear.

Now that it's fixed tho my body decided I would need to pee every night about 2 hours after I went to bed... La vieillesse est un naufrage.


At times, when my SO is snoring, I just put my overhead headphones, turn on noise cancelling, put some gentle music on and lay on my back until I fall asleep.

Not exactly great, but does the job.

Using ear plugs in the past caused infections as I have curved ear canals.


I'd never go to sleep with AirPods in intentionally.

I had a scare in the very early days of owning AirPod Pros accidentally falling asleep with them, and waking up with tinnitus that took hours to go away.

Since then with every gen I've occasionally had ANC malfunction when they come in contact with a pillow and make some deafening noises, I wouldn't be surprised if that day they had been making some horrible shrieking noise while I slept.

(in fact, as if to mock me, I left this comment unsent realizing I didn't care enough to finish it, until my AirPod Maxes decided to flashbang me for having the audacity to while lying on my back. people need to know these things are treacherous in bed)


I like the sleep ear plugs; they're silicone plugs that are moldable. So they don't tend to extend out so much, or over-fill the ear like buds do.

I was at a festival a year back and unfortunately our camp was near a minor side stage which was running a daytime talent show. I wanted to take an afternoon nap. The sleep ear plugs + noise canceling headphones dulled the moderate noise down to almost nothing.


Who cares about legality in these matter ? You are not going to see a UN white ship splitting the Detroit between Iran and Oman.

It's whoever has the will and the mean.

The more I see of that war, the more I think a war was/is necessary because we can't let these lunatics have nuclear weapon. Whatever the cost.


You make a lot of claim yet gives no source or material to back up your claim.

Beside, what would be a great distribution beyond grapheneos. iOS isn't, stock Android is much worst, calyxos ? Lineageos ? They are much worst on the security.


> Our messaging system has long been designed to balance user privacy with the ability to respond to scams, harassment, and other safety concerns when users report them or when required by law

TikTok about why they won’t put e2e for private messages.

I guess it’s reasonable to give up privacy to save the children, TikTok cares so much about our kids safety and wellbeing !


This is awful. They are doing this so they can literally advertise to kids. I bet their dbs aren't encrypted at rest either. Complete foolishnes

Can you steelman TikTok's argument?

No, because it's terrible. There's no need to break encryption to allow you to report a user. You'd just report via a copy of an excerpt of the conversation and leave the rest of your communication private. If the user can't tamper with the extraction of that excerpt, you can trust it is correct. You could even extract hashes from both the reporting party and the reported party and compare them with zero knowledge of the actual conversation.

More sophisticated HNers can chime in with zero-knowledge proofs and whatnot to show that their argument is DOA.


You have you contrive a complex system involving zero knowledge proofs and a lot more work, rather than just being able to see on the server the message asking for a child to dance in their underwear directly makes their argument DoA?

Hashing a handful of strings and comparing them is incredibly simple. Not sure where you think the complexity lies. It is in fact so trivial I'm confident Claude Code can one-shot this.

But I'll humor the notion and say that, yes, I'd rather not let them see me dancing in my underwear unless I've explicitly decided to share it.


It's pretty easy to do tho and it has the benefits of both allowing privacy and being able to check if someone is actually trying to abuse kids.

No e2e is understandable if you're the Chinese government but isn't TikTok now run by a us company (at least in America) ?


HN isn't a place for serious thought nor internal critique. No one (all bots at this point?) will critically engage past the most surface level reddit tier argument.

I'm sorry but encryption like https has been around since 1995. Every software engineer knows they should use encryption to protect pii.

What am I supposed to be doing on behalf of meta? To prove I am not a bot with a smooth brain? Present their argument? Their argument is they want to sell as much data as possible about adults and children to the department of war. They want to double dip and use all user data as a means to produce advertisements, algorithms, and user experiences that negatively impact children's health so they can profit.

https://www.abc.net.au/news/2026-03-26/meta-and-google-found...

If you want me to argue for meta on their behalf to help them find reasons to forward their goals of exploiting their user base, I won't. The exercise has negative value.


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

Search: