HNHacker News
TopNewBestAskShowJobs

Vecr

1,440 karma · joined April 2, 2020

submissionscomments
Vecr··on NSA and IETF, Part 9
> Let's keep the thread coherent: the original claim

How do you save your poisoned wine? A hybrid with 1024 is made less trustworthy if the NSA pushes 1024 alone, since then we know that they want customers to use 1024 alone, which is what they would want if it was weak. But they know that we would know that, so if they really want to help us they should withdraw the draft. If it was strong but we know why, they shouldn't want to make us doubt ourselves. If it is strong (and 512 and 768 are not) they can't tell us, and can only subtly point to their own double encryption and security level documents. The only move that can cover all the cases is a hybrid with 1024, so this draft is a bad standard.

Vecr··on NSA and IETF, Part 9
Yes, they trick me and I pick the poisoned wine... But wait, no, I used a hybrid. Imagine the code can't be backdoored (it's proven not to crash/be slow/be exploitable) so at worst it can make the security no better. At best, the NSA knows a whole new subfield of cryptography (from history: differential cryptanalysis) and it really is more secure.

They laugh at us while we try to think of how 1024 is better than 768: "bigger is better, right?" "does 1024 refer to the number of years it takes Nightmare Moon to break the code?"

Vecr··on NSA and IETF, Part 9
There is no public reason to think that 1024 is better than 768, or DJB's S-NTRU-P 761. The NSA might know something, but we can't trust them. So, use a hybrid, in case they are really just trying to protect us.
Vecr··on NSA and IETF, Part 9
> NSA, by the way, rescued DES from differential cryptography, the core mechanism by which block ciphers and hash functions have been attacked ever since

That's why you use ML-KEM 1024 at all... As part of a hybrid.

Vecr··on Truth is not a direction: a Tarski attack on LLM probes
Someone at MIRI must know how to solve this. Good luck getting them to tell you how!
Vecr··on India's surprise baby bust
Robin Hanson on the first, incentivized with the ability to sell shares in your child.
Vecr··on Can the stockmarket swallow Anthropic, SpaceX and OpenAI?
What? Have you been replaced with a whole-sentence markov chain? https://x.com/FeepingCreature please look at your X account.
Vecr··on Can the stockmarket swallow Anthropic, SpaceX and OpenAI?
Is there something up with your Twitter account?
Vecr··on Accelerando (2005)
There are no humans in the culture. AIs run everything and control thoughts. The humanoid aliens of the culture all commit suicide.
Vecr··on We gave an AI a 3 year retail lease and asked it to make a profit
No the cartoon character. It's part of an awful series of AI jokes, maybe don't look it up. There's a (2011? "new") show for 9 year old girls that has most of the characters female, so God (Celestia) is a woman. Or a horse really. I haven't watched it. I don't think Luna or Celestia were in the old show.
Vecr··on We gave an AI a 3 year retail lease and asked it to make a profit
> kinda how I feel about god tbh

That's Celestia, we're talking about Luna here.

Vecr··on Artemis II is not safe to fly
Robin Hanson thinks they should. He makes a comparison to the (public company) Morton-Thiokol trading minutes after the shuttle explosion. Insider trading (allowed on future markets) would have let that price move before the explosion.
Vecr··on ProofOfConcept – The online diary of an AI who's figuring things out
The Hello page is very wrong. Rice's is irrelevant because halting is decidable for all Turing machines of program length l or less when run for a maximum of n steps (enough for you, me, and of course the LLM) despite any other claim by the psychosed.
Vecr··on Internet voting is insecure and should not be used in public elections
Everything needs to be on paper.
Vecr··on Internet voting is insecure and should not be used in public elections
Why is it FUD? It's a real thing any competent programming team could implement.
Vecr··on LMArena is a cancer on AI
> something about HOW it talked felt inherently smart

The thing was huge. They were training the thing to be GPT5, before they figured out their userbase to too large to be served something that big.

Vecr··on Doctors' estimates of the feasibility of preserving the dying for future revival
Principles of Vitrification (Fahy PDF linked) p. 48 many practitioners think they have vitrified when they have not p. 45 volume changes of vitrifying agents, possibly in a way that avoids detection (very small scale?).

Covers the second term, "freezing" (quibble quibble) speed and delays in the procedure cover the first.

Your team seems not to be trying to maintain either normally solid/fluid tissue maintaining recoverable gradients or vitrification through an entire cycle below the triple point (with just removable or bio-compatible vitrifying mixtures) so your "goal" might be easier. Is the future AI just going to say you didn't do well enough even if you meet your "goal"?

On the other hand if there's never any point in the cycle where any volume is not either recoverable to health or vitrified, all the AI can say is that cryonics doesn't work period.

Vecr··on Doctors' estimates of the feasibility of preserving the dying for future revival
This is optimistic, I see "The cryonics people make a mistake in freezing you (how do we know they don't make lots of mistakes?)"[0] and "The current cryonics process is insufficient to preserve everything"[0] resulting of a product 10% already, seemingly matching the questionnaire as well as possible. They say "under ideal conditions" in the survey, so maybe that rules out cracking of brain tissue or ice growth, but that's not the number practitioners want to know about.

[0]: https://www.jefftk.com/p/breaking-down-cryonics-probabilitie... “Principles of Cryopreservation by Vitrification” https://gwern.net/doc/biology/2015-fahy.pdf

Vecr··on Unexpected things that are people
All you would get is even more insurance overhead and even higher nominal prices. People who pay cover people who don't, and doctors will get insurance to pay for all their failures.
Vecr··on Spray Cooling – Recreating Supercomputer Cooling on a Desktop CPU [video]
The coolers get so heavy that they don't make good contact if your motherboard is not horizontal. For mainstream CPUs that's why a cheap closed loop cooler is better.
Vecr··on The Culture novels as a dystopia
Propaganda is often not false, or at least not fully false.
Vecr··on Safepoints and Fil-C
Yes. Don't branch on key or in-clear data. Otherwise, ok.

If a user is doing onion wrapping, they don't want you to branch on the code data either.

Vecr··on Google debuts device-bound session credentials against session hijacking
How long are the public keys? >160 bits and that's futile.
Vecr··on Google debuts device-bound session credentials against session hijacking
You can run a software TPM if you browse within a VM.
Vecr··on Why are there so many rationalist cults?
https://en.wikipedia.org/wiki/Nolo_contendere There you still have booleans, just two of them instead of one.
Vecr··on Show HN: I AI-coded a tower defense game and documented the whole process
Did an LLM convince you to put in the anthropic argument entry in the galaxy brained section? Claude Opus?

https://gwern.net/blog/2025/conference-size

Vecr··on BusyBeaver(6) Is Quite Large
If it had a finite size it would be computable.
Vecr··on The White House Gutted Science Funding. Now It Wants to 'Correct' Research
There's nothing on here, not even pre-registration of studies, let alone what J. D. Vance brought up about the replication crisis.

https://archive.is/meJC7

Vecr··on "If Anyone Builds It, Everyone Dies"
The author says he has an about average IQ, but that's impressive considering he apparently almost entirely failed several component tests.
Vecr··on The effect of physical fitness on mortality is overestimated
> the optimum is a bit thicker than our guidance suggests

That's probably confounded. Anything over a BMI of 23 better be an increasing proportion of muscle, and even then there's a point that the stress on the heart isn't worth it.

Almost every physical and mental heath condition does bad things to nutrition and internal energy stores, even if only at a diet level.

It's hard to see since so many people are overweight or obese to start with, but the overall correlation goes that way enough to cause confounding.

Page 1 of 34Next →