dekhn

35 ms·

931.
by dekhn 2y ago
https://dellweb.bfa.nsf.gov/awdlst2/default.asp shows the NSF funding for Wesleyan. You can drill down and infer some of the details about the funding programs.
932.
by dekhn 2y ago
damn, I got nerdsniped again
933.
by dekhn 2y ago
Yeah, when I saw the original comment I tried to find the source of the term but wasn't able to find it. To me it sounds like medical genetics terminology (known for terms like "penetrance", "allele", "epistasi
934.
by dekhn 2y ago
dose is such a weird term for "copies"
935.
by dekhn 2y ago
Intepretability is nice, I guess, but what if the underlying latent model for a real-world system is not human-understandable. if a system provides interpretability by default, does it fail to build a model for a system that can't be
936.
by dekhn 2y ago
I knew somebody who called that "Barf dip". presumably for what it looked like, not tasted like.
937.
by dekhn 2y ago
There's a whole concept of "farmer's cheese"- quickly prepared from pressing whey, minimal preservation- intended for nearly immediate consumption. Cottage cheese, queso fresco, paneer, ricotta, are all examples... then
938.
by dekhn 2y ago
I'm pretty sure the MSG disassociates into the sodium ion and the glutamate, since it's soluble in water. I only use the term MSG to refer to the additive when it's in the salt form in a dry jar. Side note: it's really
939.
by dekhn 2y ago
Umami is a lot more present that people recognize. I've built up an intuition for this over the years, and also sort of trained my tongue. What we call umami is a subjective experience that has an underlying molecular cause, but it&#x
940.
by dekhn 2y ago
Cheese crystals are umami. Many of them are glutamate crystals. I am curious if the other amino crystals have a similar flavor profile.
941.
by dekhn 2y ago
Bonds jiggle and wiggle with characteristic times in femtoseconds, and most people believe it's not necessary to simulate at higher frequencies than that.
942.
by dekhn 2y ago
Why do you keep talking about LLMs? An AI research scientist would almost certainly contain many more components.
943.
by dekhn 2y ago
There's a single scene in the film- when the ship takes off and you see New York hundreds of years in the future- that has more detail and plot potential in it than nearly any other movie I've seen. https://www.reddit.
944.
by dekhn 2y ago
I think you must have misunderstood my statement on multiple levels (or you're being disingenuous; it's hard to tell). I don't think you can ever prove anything to be true in science, and I expressed that in my statement (&q
945.
by dekhn 2y ago
you're conflating LLMs with AI. Strictly speaking, from what we understand of physics, chemistry, biology, computing, and mathematics, there is no coherent argument that you could not build an AI which could be an effective research s
946.
by dekhn 2y ago
I can't find any source that says this. It's hard to check for sure because there are lots of articles about the student loan bailout. But I was unable to find even a single source saying this.
947.
by dekhn 2y ago
Only took ten years: https://www.youtube.com/watch?v=BpsMkLaEiOY At the time, Fitzpatrick worked for Google and got in a small amount of trouble for posting this video. The nest creator (Fadell) and the head of INfrastruct
948.
by dekhn 2y ago
Maybe 23andMe was (certainly they seemed to be based on the idea that all disease would be single-gene, single-mutation, mendelian inheritance). But many modern treatments for cancer use genetic tests to determine if the treatment would be
949.
by dekhn 2y ago
It's not that hard to find decent lenses. I bought a lens with a long focal length and wide aperture from surplusshed.com and installed it into a PVC tube, then attached my camera to it (no lens). I used it for solar observation- a f
950.
by dekhn 2y ago
yes, i read what happened (the twitter thread and the various other posts). We're mostly in alignment, I just don't see where Columbia incentivized this.
951.
by dekhn 2y ago
Yeah, there's also an likely apocryphal story about tanks and machine learning: https://gwern.net/tank The more you work with large-scale ML systems the more you develop an intuition for these kinds of properties. If
952.
by dekhn 2y ago
Projects like this are a great demonstration of just how far hobby projects can go today by exploiting some fairly recent innovations. In particular, designing a custom lens geometry and having JLC manufacture small runs cheaply is especia
953.
by dekhn 2y ago
Did you look at what the poster said they did?
954.
by dekhn 2y ago
I dunno. My perspective is that I've worked in ML for 30+ years now and over time, unsupervised clustering and direct featurization (IE, treating the image pixel as the features, rather than extracting features) have shown great utili
955.
by dekhn 2y ago
I use the term genetic history, rather than race, as race is only weakly correlated with body level phenotypes. If your question is truly in good faith (rather than a "I want to get in argument "), then my answer is: it's com
956.
by dekhn 2y ago
The title is highly misleading. The author of the article wrote a tool to "cheat" in LeetCode interviews, violated numerous interview policies, and stepped on a lot of toes. All while trying to pitch their product as both a comp
957.
by dekhn 2y ago
Nearly all radiology practice has integrated AI to some degree or another at this point.
958.
by dekhn 2y ago
It doesn't seem surprising at all. Genetic history correlates with race, and genetic history correlates with body-level phenotypes; race also correlates with socioeconomic status which correlates with body-level phenotypes. They are
959.
by dekhn 2y ago
Studies like that, no matter how careful, cannot say anything about causation.
960.
by dekhn 2y ago
My initial imposter syndrome came from not being able to understand enough machine language and the Apple II graphics format to be able to duplicate Ultima IV (with simultaneous sound and animation). But this exists now: https://

More ›