jkhdigital

16 ms·

91.
by jkhdigital 2y ago
I had the same thought. It reminds me of solving Project Euler problems, where there is often an obvious naive approach which is guaranteed to produce the correct answer but would consume prohibitive memory/compute resources to execute
92.
by jkhdigital 2y ago
All major technological advancements have eliminated some jobs and created others. In general, producing more with less human input is how society becomes richer. Is it a tragedy that we no longer need telephone switchboard operators, ice c
93.
by jkhdigital 2y ago
Here, let me drop this mic for you
94.
by jkhdigital 2y ago
The lab wasn’t just studying coronaviruses. The director had intimate knowledge of gain-of-function techniques, with publications and grant proposals to document this. Some of the research was published during her tenure at the lab, so it
95.
by jkhdigital 2y ago
You're not wrong, but natural language has a lot of stylistic "noise" which can be utilized as a subliminal channel without noticeably degrading the semantic signal.
96.
by jkhdigital 2y ago
Why do you need to rerun the LLM? Watermark detection only requires the hash functions (equation (1) from the paper).
97.
by jkhdigital 2y ago
This isn't necessarily true, it just depends on the implementation. I can say that because I've published research which embeds steganographic text into the output of GPT-2 and we had to deal with this. Running everything locally
98.
by jkhdigital 2y ago
No you've got it right. Watermarks like this are trivial to defeat, which means they are only effective against lazy users like cheating college students and job applicants.
99.
by jkhdigital 2y ago
I spent the last five years doing PhD research into steganography, with a particular focus on how to embed messages into LLM outputs. Watermarking is basically one-bit steganography. The first serious investigations into "secure"
100.
by jkhdigital 2y ago
I really don’t know why I stumbled into the comments section on this particular article, but while I’m here I have to commend you on writing perhaps the most thoughtful and eloquent comment I have ever read on HN.
101.
by jkhdigital 2y ago
There are a lot of nuances to these situations of third-party involvement and the ruling discusses these at length. If you’re interested in the precise limits of the 4th amendment you should really just read the linked document.
102.
by jkhdigital 2y ago
I’m not sure if I am mathematically sophisticated enough to follow along but I’ll try. This chain of thought reminds me of the present state of cryptography, which is built on unproven assumptions about the computational hardness of certain
103.
by jkhdigital 2y ago
About four years ago my parents got a cabin up in the mountains right along the NC-TN border, and we’ve been up there almost every summer since. Just last month we went tubing on the North Toe, so I know exactly where that CSX rail line is.
104.
Ask HN: How should I teach undergrad intro to programming?
4 points by jkhdigital 2y ago|2 comments
105.
by jkhdigital 2y ago
…did you read the link you just posted? If the person whose words are being presented as evidence is available for cross-examination then (legally) it’s not hearsay. The defendant in a criminal case is always available, so any statement t
106.
by jkhdigital 2y ago
I don’t think anyone is actually trying to connect with distant cousins for social interaction; rather the shared DNA segments provide hard evidence to confirm distant ancestors. I got into genealogy a couple years ago after thumbing throug
107.
by jkhdigital 2y ago
Location: Tokyo, Japan Remote: Yes, please Willing to relocate: No Technologies: Elixir, JS, Lua Résumé/CV: www.linkedin.com/in/jimhowes Email: jotakami (at) gmail (dot) com Midcareer transition after military and consulting
108.
by jkhdigital 3y ago
Because if it were any other way then you wouldn’t exist to sit there and ponder the question. That’s the unsatisfying answer. I think it makes sense to draw an analogy to evolution—stable arrangements of elementary particles that (somehow)
109.
by jkhdigital 3y ago
Profit-maximizing business is the engine of economic growth. Looking over the past couple hundred years, I’d say the outcomes have been nothing short of spectacular.
110.
by jkhdigital 3y ago
Sounds like he had a schizophrenic psychotic break. His behavior and especially his conspiratorial comments are classic schizophrenia. (Source: My sister and late aunt are schizophrenic)
111.
by jkhdigital 3y ago
My understanding was that a large tranche of that $10B consisted of Azure compute credits, not actual cash.
112.
by jkhdigital 3y ago
There are two use cases: (1) when the computation is proprietary so the user can’t run it, or (2) when the user doesn’t even want to trust their own hardware.
113.
by jkhdigital 3y ago
It would be hideously expensive due to the compute required and the total inability to monetize the data. Some people value their privacy that much; most don’t.
114.
by jkhdigital 3y ago
Whatever Numerai has done to their data, there is no indication that they actually transformed it in a way that would hold up to a theoretical cryptographic adversary.
115.
by jkhdigital 3y ago
Still doesn’t quite justify homomorphic encryption if the computations involved are mundane (like tax calculations). The true use case is when the computation itself is proprietary, so that users don’t want to reveal their data but the prov
116.
by jkhdigital 3y ago
There is no side-channel because FHE effectively considers the computational hardware itself to be the adversary.
117.
by jkhdigital 3y ago
It is another accident of evolution that we ended up with 12 finger joints, allowing one to count to twelve with a single hand by tapping the thumb on each finger segment in turn.
118.
by jkhdigital 3y ago
Back in 2018 I took a graduate level distributed systems course from a professor who spent most of his lecture time rambling on about how we’d be better off reading HN than our textbook. There were no exams, only four projects which involve
119.
by jkhdigital 3y ago
For all the truly deceptive and harmful propaganda that spews forth from government mouthpieces, it really boggles my mind that this is the one that “media watchdogs” decide to pounce on.
120.
by jkhdigital 3y ago
As with nearly all of social science, we can’t really run the controlled experiments necessary to determine if this is a causal relationship. Just because it looks like a “monetary problem” does not mean that throwing money at it will fix a

More ›