Viewing profile — danielmarkbruce

This is a regulatory issue. Various states have all kinds of stupid laws about stored value, escheatment, money transmission etc that kick in when you have "credits". It's quite st…

yes, it is the annoying part. I just use a wok. Since I'm only deep frying for 30 seconds and it's fully submerged in the oil, it's... not tooooo bad. But it's annoying to do and q…

yes and yes. You can build and train a model in about 15 lines of pytorch. And you can build and break your own 8 bit xor cipher in about 10 lines of python. Hacker news is full of…

Yes, indeed. And if I really had a good idea for an architecture, and millions to blow trying different ideas... I would be doing that and not on hacker news.

Many do. They don't start when you order....

For a filet for 1 hour, you can sort of do what you want. I add a tiny bit of olive oil and salt, but not much. If you start going longer times with the sous vide, you can get diff…

The discussion is about a good steak....

you really don't for most things. Ziplock is fine for things that take an hour or two.

Sure - but you shouldn't be running a ribeye, as an example, at 130. Run it at 138 for 4 hours.

Then don't go to a good restaurant. Sous vide is extremely common, and oils of various types are used liberally.

Corn oil is cheap and available at costco. A sous vide is a fancy french name for a $150 device from target that keeps water at a specific temperature... This "simple process" is m…

Go to costco, buy prime filet. Put it in a ziplock in a sous vide at 130 for an hour. Drop it in corn oil at 425 for 30 seconds. Add some butter and salt. That simple process, sugg…

Yup, sure. But my claim is that I believe it's better than 20-1 against that a model can break it. The fact that many encryption schemes have later been shown to be fragile and tha…

Yes, "magic is possible" if you defined "magic" as "very large models approximating functions in a way that people didn't think would work". Current SOTA language and vision models…

>> It won't work because these algorithms are designed to have no patterns at all Every encryption algorithm proposal has this property of being designed to have no patterns....

Ouput: 128 logits. Input: maybe 10 samples of plaintext,ciphertext (using the same key), so maybe a 2560 length tensor. Loss function: binary cross entropy on the true key bits. Ar…

I can't tell if you are serious at this point. I literally say, twice, that I'm not talking about a language model. And I also say the model would predict the key...not the plainte…

Well, again, i'm not talking about a language model.... And, just because what I'm saying isn't especially likely to work, it's not obvious that it cannot. Very large models are do…

It probably isn't. But if you laid 20-1 I'd bet a large model will break an industry used standard within 10 years. That's a loose framing of a bet, but I think you get my point, e…

Train on plaintext, ciphertext -> key.

This is kind of a stupid argument. How about make a slightly stronger claim like "models won't break symmetric crypto" ? I mean, language models aren't even trained to break symmet…

Building "simulators" that use ML/AI instead of running the calculations every step is a thing.

in AI/ML, no. They are just going to automate AI/ML research to start with. Totally doable. For some of the other things, undoubtably yes.

Automating ML/AI research seems completely tractable. Most of the other claims seem much less doable.

While this is right on some level... the incentives aren't the main driver. Finding good targets in biotech is just a harder problem to solve. Like way harder. Disease biology is c…