Veedrac

91 ms·

931.
by Veedrac 7y ago
Just buy a Gameboy. Good games are timeless.
932.
by Veedrac 7y ago
If you're willing to overlook that Z3 is a more general SMT solver, it's a useful tool for one-off problems where raw brute-force is inefficient. I used this to proof-of-concept an attack on Seahash[1]; this was the first time I&#
933.
Is AI Safety a Pascal's Mugging?(youtube.com)
1 points by Veedrac 7y ago|0 comments
934.
by Veedrac 7y ago
Thank you, I've edited the article to cite this.
935.
by Veedrac 7y ago
I pretty much agree with you: it's hard to program, but possibly something we just need to do anyway. It's a much more efficient way to build computers, and the relative benefits only look to get bigger... but if I knew how to mak
936.
by Veedrac 7y ago
I am reading the comments, and am open to critique or general suggestions. “I found this part confusing” is also very welcome. This is quite an information dense work, and a little rough around the edges, though I am pretty happy with it ov
937.
To reinvent the processor(medium.com)
223 points by Veedrac 7y ago|87 comments
938.
by Veedrac 8y ago
The paper does mention RDMA.
939.
by Veedrac 8y ago
I mostly agree, but with two caveats. We do actually know that ‘neural networks’ are poor models for the brain. Backpropagation is unphysical and brains aren't acyclic graphs, for example. What there's uncertainty about is the deg
940.
by Veedrac 8y ago
Neuro-morphic, having the form of neurons. Recently the buzz refers to circuits that perform neural network operations (the AI kind) using hardware properties directly, storing weights and performing inference or backprop with the computing
941.
by Veedrac 8y ago
To those who haven't yet read the article, don't bother. Here's a 15-second rendition of the argument to save you the time: A rat study showed most of a rat brain was used to process information from touch. This does not matc
942.
by Veedrac 8y ago
Given communication is point-to-point, and very coherent, I'd think this is extremely easy to solve if glare in the system was a real issue. Aim the receiver through a tube (or array of tubes) painted black, for instance. Certainly eas
943.
by Veedrac 8y ago
The sun is pretty small compared to space, something like .0005% of the full sky sphere, and almost half that time it's covered by Earth. Since there's no atmospheric glare in space, I suspect it's not a problem given decent
944.
by Veedrac 8y ago
Accuracy is not certainty. The one that's most informative is the one that's demonstrating greater intelligence.
945.
by Veedrac 8y ago
`casefold` is for caseless comparisons, `lower` is for display.
946.
by Veedrac 8y ago
The standard term seems to be ‘permutable’, for A(B(x)) = B(A(x)). One can also say these functions commute with respect to composition. This is slightly different from the classical notion of commutativity, since the operator (composition)
947.
by Veedrac 8y ago
This is the standard academic philosophy viewpoint, but it makes no sense. What you need are principled people with a vested interest in intellectual honesty and making a real difference. Progress simply isn't made by optimizing for di
948.
by Veedrac 8y ago
It's unfortunate that HN is more willing to downvote dissenting ideas like this than discuss them. Yes, the loss of privacy the internet has brought has been immense. However, the complete overhaul of social structures, the influence o
949.
by Veedrac 8y ago
This paper discusses a major issue that I think deserves way more attention from the academically literate crowd. This is the kind of knowledge that should be taught in secondary school, at least in the modern age with papers at everyone&#x
950.
Statistically controlling for confounding constructs is hard (2016)(journals.plos.org)
143 points by Veedrac 8y ago|37 comments
951.
by Veedrac 8y ago
> The ability for PJPEGs to offer low-resolution ‘previews’ of an image as it loads improves perceived performance – users can feel like the image is loading faster compared to adaptive images. My understanding was that progressive JPEG
952.
by Veedrac 8y ago
It's worth remembering that the data Google collects does actually influence your searches; it might not be you per se.
953.
Considerateness in OpenAI LP Debate(medium.com)
2 points by Veedrac 8y ago|0 comments
954.
by Veedrac 8y ago
As far as I can disentangle from [1], OpenAI posits only moderate superhuman performance. The profit would come from a variant where OpenAI subsumes much but not all of the economy and does not bring things to post-scarcity. The nonprofit w
955.
by Veedrac 8y ago
No, I imagine it's fairly simple to determine that humans use nonlocal context in image recognition.
956.
by Veedrac 8y ago
I consider the paper “ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness” essentially mandatory follow-up for anyone interested in this. It goes further with its conclusions and introduc
957.
by Veedrac 8y ago
Thanks so much, this is it! The play online link was what I'd been linked to. https://iplayif.com/?story=http%3A%2F%2Fwww.ifarchive.org%2F...
958.
by Veedrac 8y ago
I once found (plausibly from another HN commenter) a text based adventure where (almost?) all the words used were replaced with alternative English-sounding nonsense words, but have never rediscovered the link. I feel this would be of inter
959.
by Veedrac 8y ago
ThunderX2 isn't made by Apple.
960.
by Veedrac 8y ago
> For one, the tabloid could argue it was merely trying in good faith to resolve a dispute with Bezos—not attempting to blackmail or extort him. This reasoning is why, for example, a lawyer can send someone a letter saying their client w

More ›